Monetary Policy & the Economy

D:HI:GG:>8=>H8=:C6I>DC6A76C@ :JGDHNHI:B

MONETARY POLICY & THE ECONOMY Quarterly Review of Economic Policy

Two hundred years of central banking in Austria: selected topics

Stability and Security. Q3– Q4/16 Monetary Policy & the Economy provides analyses and studies on central banking and economic policy topics and is published at quarterly intervals.

Publisher and editor Oesterreichische Nationalbank Otto-Wagner-Platz 3, 1090 Vienna, Austria PO Box 61, 1011 Vienna, Austria www.oenb.at [email protected] Phone: (+43-1) 40420-6666 Fax: (+43-1) 40420-046698

Editorial board Ernest Gnan, Doris Ritzberger-Grünwald, Helene Schuberth, Martin Summer

Managing editors Clemens Jobst, Claudia Kwapil

Editing Douglas Deitemyer, Graeme High, Rena Mühldorf, Irene Popenberger, Arwen Ross, Rita Schwarz, Susanne Steinacher

Translations Susan Starling, Susanne Steinacher

Layout and typesetting Walter Grosser, Andreas Kulleschitz, Robert Musil, Melanie Schuhmacher

Design Information Management and Services Division

Printing and production Oesterreichische Nationalbank, 1090 Vienna

DVR 0031577

ISSN 2309–3323 (online)

May be reproduced for noncommercial, educational and scientific purposes provided that the source is acknowledged.

Printed according to the Austrian Ecolabel guideline for printed matter.

Please collect used paper for recycling. EU Ecolabel: AT/028/024 Contents

Call for applications: Visiting Research Program 4

In focus: Two hundred years of central banking in Austria A (not so brief) history of inflation in Austria 6 Christian Beer, Ernest Gnan, Maria Teresa Valderrama

The measurement of inflation in Austria: a historical overview 33 Manfred Fluch

Two centuries of currency policy in Austria 61 Heinz Handler

The financial relations between the Nationalbank and the government 77 Doris Prammer, Lukas Reiss, Walpurga Köhler-Töglhofer

Florin, crown, schilling and euro: an overview of 200 years of cash in Austria 94 Clemens Jobst, Helmut Stix

Cashless payments in Austria: the role of the central bank 120 Hans Kernbauer

Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931 140 Clemens Jobst, Kilian Rieder

The changing role of macroprudential policy in Austria after World War II 163 Sophia Döme, Stefan W. Schmitz, Katharina Steiner, Eva Ubl

The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator 190 Stefan W. Schmitz

Notes List of studies published in Monetary Policy & the Economy 212 Periodical publications 214 Addresses 216

Opinions expressed by the authors of studies do not necessarily reflect the official viewpoint of the Oesterreichische Nationalbank or of the Eurosystem.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 3 Call for applications: Visiting Research Program

The Oesterreichische Nationalbank to the department’s computer re- (OeNB) invites applications from exter- sources. Their research output may be nal researchers (EU or Swiss nationals) published in one of the department’s for participation in a Visiting Research publication outlets or as an OeNB Program established by the OeNB’s Eco- Working Paper. Research visits should nomic Analysis and Research Depart- ideally last between three and six ment. The purpose of this program is months, but timing is flexible. to enhance cooperation with members Applications (in English) should of academic and research institutions include (preferably postdoc) who work in the • a curriculum vitae, fields of macroeconomics, international • a research proposal that motivates economics or financial economics and/ and clearly describes the envisaged or pursue a regional focus on Central, research project, Eastern and Southeastern Europe. • an indication of the period envis- The OeNB offers a stimulating and aged for the research visit, and professional research environment in • information on previous scientific close proximity to the policymaking work. process. Visiting researchers are ex- Applications for 2017 should be pected to collaborate with the OeNB’s e-mailed to research staff on a prespecified topic [email protected] and to participate actively in the by November 1, 2016. department’s internal seminars and Applicants will be notified of the other research activities. They will be jury’s decision by mid-December. The provided with accommodation on following round of applications will demand and will, as a rule, have access close on May 1, 2017.

4 OESTERREICHISCHE NATIONALBANK In focus: Two hundred years of central banking in Austria A (not so brief) history of inflation in Austria

Christian Beer, We analyze the eventful history of inflation in Austria over the past two centuries against Ernest Gnan, the background of institutional, economic and political developments. Ultra-high inflation Maria Teresa or hyperinflation resulted from repeated monetary financing of war expenses, war-related Valderrama1 destruction of productive capacity as well as fear of short-term political and social consequences of anti-inflationary measures. Like other countries, Austria experienced the highest inflation during peaceful times in the 20th century when efforts were made to cushion the negative output effects of oil price shocks. Inflation volatility was high in the 19th century up to the Gründerzeit boom and in the 20th century during the Great Inflation; it was low in the half-century up to World War I and during the Great Moderation. Our frequency domain analysis for the pre-World War I and post-World War II periods finds that, in line with the economic literature, the correlation between money growth and inflation is considerably higher for long and very long frequencies than for business cycle frequencies. The varying correlation between money and inflation reflects changing monetary regimes. We cannot establish a stable empirical Phillips curve relationship; in the post-World War II period, the relationship breaks down once supply shocks (oil prices) are included. Deflation was quite frequent in 19th-century Austria and was not necessarily associated with recessions. By contrast, Austria’s 20th century Great Depression fits the textbook notion of deflation, high unemployment and economic slack. Formal central bank independence did not prevent the erosion of monetary value in “emergency situations” of wars. Hyperinflations and currency reforms repeatedly inflicted substantial and long-lasting damage on citizens’ trust in state money and in the state more generally. JEL classification: E58, N13, N14 Keywords: inflation, Austria

Maintaining price stability is the pri- ing economic circumstances (such as mary objective of monetary policy the conduct of devastating and expen- nowadays, and we have become used to sive wars) and the evolution of economic very low rates of consumer price infla- thought (e.g. the doctrine of metal tion in Austria over the past three standards that applied throughout most decades. However, inflation has not of the 19th century, the Keynesian con- always been low. The bicentennial of cept of exploiting an – actual or per- the Oesterreichische Nationalbank ceived – tradeoff between unemploy- (OeNB) offers the opportunity to ana- ment and price stability during the lyze the eventful – and at times dra- 1960s and 1970s, and the emphasis on matic – history of inflation in Austria time inconsistency and central bank over the past two centuries. independence since the 1980s). Looking back in history, we find Any analysis of price and inflation that the central bank’s role, mandate, developments should take this broader and institutional status in relation to perspective into account rather than the government, the intermediate goals imposing present-day economic and of monetary policy as well as the in- institutional concepts ex post. In this struments available to, and used by, the article, we therefore combine a historic central bank have evolved considerably narrative that recalls relevant political, over time. These changes reflect vary- institutional and economic develop- Refereed by: Geoffrey Wood, 1 Cass Business Oesterreichische Nationalbank, Economic Analysis Division, [email protected], [email protected], [email protected]. The views expressed in this paper are exclusively those of the authors and do not School and necessarily reflect those of the OeNB or the Eurosystem. The authors acknowledge helpful comments and sugges- University of tions by Walpurga Köhler-Töglhofer, Clemens Jobst, Lukas Reiss, Fabio Rumler, Alfred Stiglbauer, our referee and Buckingham participants in a workshop among the authors of this volume.

6 OESTERREICHISCHE NATIONALBANK A (not so brief) history of inflation in Austria

ments when identifying various infla- sistent policies resulting from the quest tion regimes with empirical statistical to achieve credibility for the central and econometric analysis to shed light bank and paper money despite rulers on Austria’s inflation history from vari- seeking ways to have the central bank ous angles.2 We use the long time series finance their expenses, most notably that have now been made available by for wars. Various institutional changes the OeNB to discuss specific features of had been attempted (Jobst and Ker- the development of the inflation rate. nbauer, 2016) to enhance the prede- In particular, we investigate inflation cessor institutions’ credibility by bol- volatility, the long-run relationship stering these institutions’ perceived – between money growth and inflation, rather than actual – independence. But and the short-run correlation between despite this conflict of goals (monetary the output gap and inflation in Austria; stability versus war financing) currency we then discuss whether these relation- stability was not seriously endangered ships have changed over time. until the Napoleonic wars, as taxation For our purposes, it is useful to dis- and loans sufficed to finance most war tinguish five periods of inflation in expenditure. In the end, however, the Austria over the past two centuries: Napoleonic wars proved to be too long; first, the pre-World War I period of an the pressure for the monetary financ- overall relatively stable price level (sec- ing of war expenses prevailed, and the tion 1); second, World War I (WW I) resulting expansion of money in cir- and the following period of hyperinfla- culation eroded the purchasing power tion (section 2); third, the post-WW I of money. Paper money, originally ad- currency reform, the Creditanstalt cri- opted by the public voluntarily for its sis and the Great Depression (sec- ease of use, was eventually enforced by tion 3); fourth, World War II (WW II) law, by being granted the status of legal and the postwar inflation and currency tender. Repeated attempts at currency reform (section 4); fifth, the further stabilization were half-hearted and had postwar period, which includes the re- no lasting impact. The amount of paper sponse to the demise of the Bretton money in circulation increased 120- Woods system, the first and second oil fold between 1792 and 1816. In the price shocks as well as the hard cur- two decades before the Nationalbank’s rency policy and Austria’s membership foundation in 1816, the average price of in Economic and Monetary Union seven food items in Vienna increased (EMU) (section 5). Section 6 summa- 50-fold. The sovereign default and cur- rizes and concludes. rency reform of 1811, which had at its core the compulsory devaluation of pa- 1 The first century until World per florins to one-fifth of their value, War I was considered “a brutal and wrongful The foundation of the privilegirte oes- method of debt relief that shook finan- terreichische National-Bank in 1816 re- cial and legal relationships to the core” flected the attempt to establish mone- (Brandt, 1978, as quoted and translated tary stability after 100 years of incon- in Jobst and Kernbauer, 2016). At the

2 For a detailed account of the broader economic historical and institutional developments, please refer to Jobst and Kernbauer (2016) as well as to Antonowicz et al. (2016). The latter publication also includes an overview chart of inflation in Austria spanning the period from 1800 to 2015 and numerous overview tables with relevant historical events.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 7 A (not so brief) history of inflation in Austria

Chart 1 The ﬁrst century until World War I (1816 to 1913) Price index Inﬂation Index (1914=100, logarithmic scale) % Standard deviation in % 1,000 15 8

10 6 5

0 100 4 –5

–10 2 –15

10 –20 0 1816 1826 1836 1846 1856 1866 1876 1886 1896 1906 1816 1826 1836 1846 1856 1866 1876 1886 1896 1906 Price index Inﬂation rate (left-hand scale) Inﬂation volatility (right-hand scale)

Nominal GDP and money supply Industrial production ATS billion (converted from original currency Million currency then in force Index (1913=100, logarithmic scale) % (logarithmic scale) (logarithmic scale) 1.0 100,000 1,000 30

10 0.1 10,000 100 0

–10

0.0 1,000 10 –20 1816 1826 1836 1846 1856 1866 1876 1886 1896 1906 1816 1826 1836 1846 1856 1866 1876 1886 1896 1906 Currency in circulation (left-hand scale) Index (left-hand scale) M1 (left-hand scale) Growth rate (right-hand scale) Nominal GDP 1 (right-hand scale) Nominal GDP 2 (right-hand scale) Source: OeNB, Komlos (1986), Mühlpeck et al. (1979), Kausel (1979), Bank of Greece et al. (eds.) (2014), authors’ calculations. For further information on the data, see annex. Note: Nominal GDP 1: Kausel (1979), Nominal GDP 2: Bank of Greece et al. (eds.) (2014). Before 1900 Gulden österreichischer Währung (Austrian ﬂorin), 1900 conversion to crown.

same time, the currency reform of 1811 privately owned institution was shaped did not bring the promised end to inflation. to pursue this objective. It is against this background that the However, subsequent changes in foundation of the Nationalbank in 1816 the central bank statutes successively needs to be seen. The currency reform increased the state’s influence on the of 1816 was based on three pillars: first, Nationalbank’s policy, and a coalition the return to convertibility into metal of interests of the state, the National- currency; second, the credible consol- bank’s shareholders and management, idation of state finances; and, third, who all sought to maximize seignor- the foundation of a new institution in age from note issuance, implied that charge of monetary stability, whose in- the share of banknotes in cash in circu- centive structure as an independent, lation increased to much higher levels

8 OESTERREICHISCHE NATIONALBANK A (not so brief) history of inflation in Austria

than e.g. in France, England or Germany of uncertainty, distrust and panic in the and that the Nationalbank’s silver re- run-up to the revolution of 1848. A run serves remained lower than originally on banks and on the central bank in planned. A low silver coverage ratio February 1848 prompted the National of banknotes in circulation either trig- bank to start publishing its balance gered silver purchases (1830, 1840) sheet on a monthly basis. However, this or prompted the Nationalbank to use made visible the large share of govern- moral suasion to prevail upon the few ment debt in the Nationalbank’s assets, large borrowers to curb the volume of further eroding rather than restoring discount (1836, 1840), reflecting the trust and in turn precipitating an end Nationalbank’s focus on policy mea- to the convertibility of banknotes into sures aimed at influencing the quantity silver. Attempts to restore convert- rather than the price of money. The in- ibility failed several times until 1867, terest rate played quite a minor role as mainly as a result of repeated war-in- a policy instrument; rate cuts in 1829 duced state recourse to central bank and 1833 were targeted at smoothing financing or of the issue by the state the issuance of the government’s capital of parallel paper currency of the same market financing. At the time, financ- value as the central bank’s banknotes. ing the government was still an im- This measure eroded the silver cover- portant activity of the central bank, as age ratio of banknotes in circulation. the predominance of claims against the The three years up to 1866 witnessed government among the Nationalbank’s a marked reduction in banknotes in balance sheet assets clearly showed. circulation.4 There is no agreed view on While annual inflation fluctu- whether this “deflationary” monetary ated substantially (between –7.1% and policy was responsible for the contrac- +7.5%), the level of consumer prices tion and subsequent stagnation of in- remained quite stable overall between dustrial production or whether the fall 1825 and 1844, with inflation averag- in the money supply was endogenous ing 0.5% and the cumulated increase instead, reflecting low money demand in the consumer price level coming to in the face of foreign and domestic real 5%. Inflation volatility3 also dropped shocks to aggregate demand. over this period, starting from quite Empirically, the correlation between high levels after the Napoleonic wars. growth in currency in circulation and In the view of Jobst and Kernbauer inflation during this period was relatively (2016), this apparent monetary stability high (see chart 2 and the annex for data rested on secrecy with respect to the and technical details). This is also true Nationalbank’s fragile balance sheet and for various filter frequencies when one business model as well as on the state’s of the series is lagged. A possible inter- assistance in ensuring silver convert- pretation might be that this period was ibility during periods of tension. How- without major innovations in payments ever, this very secrecy became a source systems (Kernbauer, 2016).

3 Inflation volatility is calculated as the standard deviation of the inflation rate over 11 years. We assign the result to the midpoint of the period. The length of the period chosen is both long enough for meaningful computations and short enough to capture the impact of contemporaneous structural, institutional and other historic developments. 4 Technically, this reduction of banknotes in circulation was implemented by the government reimbursing its debt with the central bank.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 9 A (not so brief) history of inflation in Austria

Chart 2 Largest correlation between inﬂation and currency in circulation or M1 Correlation coeﬃcient 1.0

0.8 6 –6 –1 0.6 –6 3 1 0.4 1 0 0.2

0.0

–0.2 –4 –1 –0.4 0 2 –0.6

–0.8

–1.0 1830 to 1866 1867 to 1913 1830 to 1866 1867 to 1913 1830 to 1866 1867 to 1913 1830 to 1866 1867 to 1913 Inflation and currency Inflation and currency Inflation and M1 Inflation and M1 (+lags/–leads) in circulation in circulation (+lags/–leads) No filter 2 to 8 years 8 to 20 years 20 to 40 years

Source: Authors’ calculations. Note: Figures attached to bars show lags(+)/leads(–) in years.

Jobst and Kernbauer (2016) charac- so-called Gründerzeit boom. The exces- terized the period between 1866 and sively expansionary monetary policy 1914 as the switch – with obstacles – was caused by a combination of sev- from fiscal to monetary dominance. eral factors: first, the continued exis- Two conditions were put into place to tence of a large stock of parallel state enable this switch: First, the new stat- paper money which was not duly taken utes of 1863 stated as the central bank’s into account in the statutory banknote policy goal pegging the value of the coverage ceiling; second, the central currency to the price of silver; given bank’s revenue motive, which pro- the stable relation of the prices for silver vided an incentive to continue its ag- and gold, this implied stable exchange gressive credit policy (since the gener- rates vis-à-vis not only the many other ation of credit increased the profits of silver and bimetallic currencies but also the central bank’s shareholders); third, vis-à-vis the pound sterling – then the the concern that monetary restriction only major gold-based currency. Second, might hurt the financial sector and the the central bank had the instruments to view that monetary policy was neither actually pursue its goal: It set the inter- able to nor supposed to prevent or curb est rate independently and it was in speculative bubbles; fourth, capital in- control of its balance sheet, thus ensur- flows from foreign investors seeking ing that banknotes were indeed backed to participate in the Austrian stock by silver, gold and private credit, rather market boom, which further exacer- than by claims against the government. bated surplus liquidity. The Gründer- Fueled by abundant liquidity and zeit boom was mirrored in a tempo- historically low interest rates, the years rary surge in consumer price inflation between 1866 and 1873 witnessed the caused by buoyant aggregate demand,

10 OESTERREICHISCHE NATIONALBANK A (not so brief) history of inflation in Austria

notably in the construction, metals and are relatively low. In fact, without lagging machinery sectors. The boom ended the variables and for both the short- and the stock market crashed in May and long-run frequencies, correlations 1873. During the subsequent two de- between inflation and money growth cades until the early 1890s, inflation was were the lowest among all the subperiods negative in two-thirds of all years, mea- for which we investigated these rela- suring around ‒1% on average. There tionships (chart 2). Possible interpreta- is no agreed view whether this defla- tions may include, first, that part of the tionary period was also accompanied Gründerzeit inflation found its way into by a depression of the real economy asset price rather than consumer price (Jobst and Kernbauer, 2016). inflation; and second, that the long pe- The Compromise of 1867, which riod of post-Gründerzeit negative infla- turned the Austrian Empire into a tion may have affected the relationship dual monarchy, implied for the central between money growth and inflation. bank that it was transformed into the To sum it up, the first 100 years Oesterreichisch-ungarische Bank. While up to WW I witnessed quite substan- renewals of the central bank’s statutes tial fluctuations in the annual rate of (now required every ten years) implied inflation, reflecting numerous shocks continuous complex political negotia- affecting aggregate demand and supply, tions for the Nationalbank, which for- such as wars and revolutions, harvest- mally reduced its independence, its dependent fluctuations in food prices, position as a single monetary entity but also crises of confidence in the cen- between two political powers which tral bank and the currency regime as often blocked each other in important well as repeated attempts by the gov- decisions implied a substantial degree ernment and central bank to re-estab- of de facto autonomy. As a result, the lish trust and stability. The main epi- Nationalbank successfully managed to sodes of inflation were associated with ensure monetary stability over the last crop failures or monetary financing of two decades before WW I and pegged wars around mid-century as well as the Austrian currency to gold, with the Gründerzeit bubble in the run-up to inflation emerging from negative terri- the 1873 crash. Periods of deflation tory in the second half of the 1890s and occurred most notably in the decade then remaining in moderate positive following the end of the Napoleonic territory between 1% and 4% during wars and – more mildly but for an ex- most years until 1913. tended period of time – after 1873. This is also reflected in the quite At the same time, the period was low inflation volatility over the 1867 to also characterized by a remarkable de- 1913 period. Apart from monetary gree of long-term stability in the level of policy and quite stable money growth, consumer prices. This was likely due to closer trade integration and technical the overall conception that monetary progress that facilitated transport and policy should in principle ensure con- communication might also have convertibility into a precious metal (mostly tributed to the downtrend in inflation silver in 19th-century Austria) at a given volatility. rate. Despite repeated recourse by the Compared to the pre-1967 period, state to central bank financing, this during 1867 to 1913 the correlations principal notion was not challenged, between inflation and the growth in and the state in several cases made currency in circulation as well as M1 efforts and took active measures to put

MONETARY POLICY & THE ECONOMY Q3– Q4/16 11 A (not so brief) history of inflation in Austria

Chart 3 Inﬂation rate from 1816 to 1908: moving averages % 6

–2

–4

–6 1816 1821 1826 1831 1836 1841 1846 1851 1856 1861 1866 1871 1876 1881 1886 1891 1896 1901 1906 United Kingdom Austria

Source: Bank of England, OeNB. Note: 11-year moving averages.

the central bank into a position to Given the ups and downs of eco- re-establish silver convertibility. As a nomic growth and inflation in 19th-cen- result, the value of money was quite tury Austria, the question arises as to firmly anchored in a long-run perspec- whether there was a short-term cor- tive. relation between economic activity and Chart 3 suggests that international inflation, i.e., to put it in 20th-century developments also played a role in de- economic terminology, whether there termining inflation in 19th-century was a Phillips curve-type relationship. Austria. Assuming that inflation in the Chart 4 displays the coefficient for the U.K. mirrored price developments of output gap from rolling regressions of commodities and other goods in the the inflation rate on the output gap world market, chart 3 suggests that and lagged inflation.5 The chart sug- global factors could explain a sizeable gests that the correlation between the portion of Austrian inflation fluctua- output gap and inflation was more pro- tions during the 19th century. Never- nounced until about the 1870s than in theless, the large positive inflation dif- the final decades of the monarchy. The ferential in the middle of the 19th cen- regression results (table A1 in the an- tury likely reflected domestic factors, nex) for the two subperiods from 1830 notably the chronic deficit in govern- to 1866 and from 1867 to 1913 also ment finances, which led to a lack of confirm this result: The estimated im- discipline in controlling the amount of pact of the output gap on inflation was banknotes and/or state paper currency statistically significantly different from circulated in parallel, thus compromis- zero only in the years from 1830 to ing repeated attempts to re-establish 1866 but not so from 1867 to 1913. silver convertibility (Jobst and Kern- Clearly, this simple analysis can provide bauer, 2016, for details). only some preliminary clues and cannot be taken as the final word.

5 See the annex for more information on method and data used.

12 OESTERREICHISCHE NATIONALBANK A (not so brief) history of inflation in Austria

Chart 4 Phillips curve from 1840 to 1903 Regression coeﬃcient on output gap 1.4

1.2

1.0

0.8

0.6

0.4

0.2

0.0

–0.2

–0.4

–0.6 1840 1845 1850 1855 1860 1865 1870 1875 1880 1885 1890 1895 1900 Coeﬃcient Coeﬃcient ±2 standard errors

Source: Authors’ calculations based on OeNB data. For further information on the data, see annex.

2 World War I and hyperinflation banknotes in circulation as well as the WW I boosted inflation and eroded the prohibition of central bank credit to the value of the currency through several state. The amount of banknotes in cir- channels. First, the war led to a massive culation increased 12-fold from mid- reduction of the labor force as well as 1914 to the end of WW I. With some the destruction of physical capital, re- lag, the escalating volume of central ducing potential output: By 1918, real bank credit to the state also resulted in GDP had dropped to less than 60% of hyperinflation. Between 1915 and the prewar level.6 Second, war expen- 1918, annual inflation averaged 84%; ditures absorbed a substantial part of in 1918, the level of consumer prices production: At 80% to 90% of prewar was 11 times higher than in 1914, which GDP, the total cumulative financial implies a striking co-movement of cost of WW I for Austria was consider- money in circulation and the price able (though relatively moderate com- level. Annual consumer price inflation pared to other countries). Third, while escalated further after the end of the WW I was mostly financed through war. In 1919, it reached 149% and in war bonds actively marketed to the 1920 99%, to skyrocket to 205% in public, one-third of financing came 1921 and to 2,877% in 1922. All in all, through direct central bank credit to between 1914 and 1922, the level of the government. consumer prices increased more than In August 1914, important parts of 5,000-fold, and by 1924, prices had the Nationalbank’s statutes were re- surged to nearly 14,000 times their scinded, abolishing inter alia the statu- 1914 level. tory minimum gold coverage of

6 Given that aggregate demand was boosted aggressively by extreme monetary expansion, most of this drop can be assumed to reflect a decline in productive capacity.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 13 A (not so brief) history of inflation in Austria

Chart 5 World War I and hyperinﬂation (1914 to 1924) Price index Inﬂation rate Index (1914=100, logarithmic scale) % 10,000,000 250

1,000,000 200 150 100,000 100 10,000 1922: 2,877% 50

1,000 0

100 –50 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 Price index Inﬂation rate

Money supply Real GDP ATS billion (converted from original currency; (logarithmic scale) Index (1830=100, logarithmic scale) % 10.0 1,000 15

0 1.0 –5

–10

–15

0.1 100 –20 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 Currency in circulation Index (left-hand scale) M1 Growth rate (right-hand scale)

Source: OeNB. For further information on the data, see annex.

Like during WW I, the central tion, albeit a few years later. While bank perceived itself as being the gov- alleviating the government’s debt bur- ernment’s printing press, causing the den, hyperinflation at the same time war-induced very high inflation to es- implied the impoverishment of the calate into outright hyperinflation. The holders of war bonds, and thus of the central bank’s management was appar- middle class. ently aware of the likely medium-run outcome of this course of action but 3 Currency reform, Creditanstalt nevertheless chose to continue on its crisis and the Great Depression path toward hyperinflation out of the Hyperinflation was ended through a perceived immediate necessity to do so, combination of measures, the first of for fear of upheaval, social chaos and which was an international rescue pro- anarchy (Kernbauer, 1995). In more gram of the League of Nations. This general terms, short-term political ne- program included guarantees by several cessities were taken as a justification for foreign states for the issuance of inter- ultra-expansionary monetary policies national bonds by Austria. This inter- that must have been – and were – national help was tied to the implemen known to ultimately ruin the monetary tation of strict monetary and fiscal order and to lead to chaos and destruc- reforms as well as close international

14 OESTERREICHISCHE NATIONALBANK A (not so brief) history of inflation in Austria

Chart 6 Currency reform, Creditanstalt crisis and the Great Depression (1925 to 1937) Price index Inﬂation rate Index (1925=100, logarithmic scale) % 1,000 10

100 2

–2

–4

10 –6 1925 1927 1929 1931 1933 1935 1937 1925 1927 1929 1931 1933 1935 1937 Price index Inﬂation rate

Nominal GDP and money supply Real GDP ATS billion (logarithmic scale) EUR million (logarithmic scale) Logarithmic scale (1830=100) % 10.0 1,000 1,000 10

0 1.0 –5

–10

0.1 100 100 –15 1925 1927 1929 1931 1933 1935 1937 1925 1927 1929 1931 1933 1935 1937 Currency in circulation (left-hand scale) Index (left-hand scale) M1 (left-hand scale) Growth rate (right-hand scale) Nominal GDP (right-hand scale)

Source: OeNB. For further information on the data, see annex. monitoring of Austrian economic poli- gram as well as by boosting official cies. Second, the Oesterreichische reserves through a League of Nations Nationalbank (OeNB) was re-estab- loan. Finally, the Schilling Conversion lished as a stock corporation in which Act, passed on December 20, 1924, the central government’s influence was provided for the conversion of crowns fairly limited and monetary financing into a new currency, the schilling, at a was strictly prohibited. Moreover, the rate of 10,000 paper crowns, or 0.694 central bank was given the mandate to gold crowns, to the schilling. Thus, the prepare a return to the pre-WW I gold schilling’s fine gold weight was fixed at standard. Third, the exchange rate of 0.21172086 g, some 30% less than the the Austrian crown was stabilized weight of the gold crown. against the U.S. dollar by ensuring the Backed by the recovery of the world credibility of the overall reform pro- economy, real GDP grew on average by

MONETARY POLICY & THE ECONOMY Q3– Q4/16 15 A (not so brief) history of inflation in Austria

3.5% per year and the budget was more 4 World War II, postwar inflation or less balanced between 1925 and and currency reform 1929. In this supportive environment The Anschluss of Austria to Germany was inflation was relatively stable. The in- swiftly followed by the liquidation of flation rate dropped strongly in 1926 the OeNB, and the German reichsmark and then remained at a moderate level replaced the schilling as legal tender. of 2.6% on average. However, despite Schillings were converted into reichsmark this generally favorable situation unem- at a rate of 1.5 to 1. Although production ployment remained high. The strong expanded sharply after the Anschluss – increase in unemployment in the early Austria’s GDP grew by 13% both in 1920s (amongst others because of a re- 1938 and 1939 – the price level, which duction in the number of public em- had stagnated during the three previous ployees by 80,000) was not signifi- years, shrank by 1% in both years. Prices cantly reversed. As a consequence, fell because the schilling-reichsmark Austria recorded an unemployment conversion rate implied an apprecia- rate of 12.7% in 1929. tion, railway ticket prices and postal The Great Depression triggered the fees were cut, the sales tax taken over breakdown of the Austrian bank Credit from Germany was lower, and wage- anstalt in 1931, leading to costly rescue price stops were enforced as well as operations by the central bank and the the German system of price regulation government. The crisis prompted the (Jobst and Kernbauer, 2016). central bank to defend the schilling The German Reichsbank financed exchange rate through interest rate the escalating war expenses of the Third hikes and foreign exchange controls. In Reich, during the last years of the war line with many other countries, Austria mostly by extending direct central ultimately abandoned the gold/foreign bank credits. In Austria, the produc- exchange parity and depreciated the tion of nonwar-related goods shrank, schilling. Austrian real GDP shrank by causing severe shortages of daily goods 22% in total in the four years from and food; however, the price freeze 1930 to 1933 and recovered only half of effective until the end of the WW II this loss in the subsequent four years up prevented inflation from breaking out to 1937. The unemployment rate in openly: Between 1940 and 1944, in- 1937 stood at 22%, indicating large and flation in Austria ranged between 0% persistent economic slack. Monetary and 2%, averaging 1.2%. By 1945, the policy during this period followed a volume of banknotes in circulation in restrictive stance, which is exemplified Austria is estimated to have increased by the decision in 1936 not to join a by a factor of 12 to 23 compared to number of European countries’ ex- 1937. As in Germany, GDP in 1945 is change rate devaluations. Inflation estimated to have halved compared to reflected the deep depression, plunging 1937. Thus, like WW I, WW II caused to –4.6% in 1931 and posting an aver- potential output to drop sharply and age of –0.3% between 1932 and 1937. money supply to increase massively, The entire period of very low inflation which implied a dramatic explosion and rising unemployment between in the price level unless measures to 1924 and 1937 suggests the existence eliminate the monetary overhang were of a Phillips curve-type relationship in taken (Jobst and Kernbauer, 2016). interwar Austria.

16 OESTERREICHISCHE NATIONALBANK A (not so brief) history of inflation in Austria

Chart 7 World War II, post war inﬂation and currency reform from 1938 to 1954 Price index Inﬂation rate Index (1925=100, logarithmic scale) % 1,000 100

100 40

10 –20 1938 1940 1942 1944 1946 1948 1950 1952 1954 1938 1940 1942 1944 1946 1948 1950 1952 1954 Price index Inﬂation rate

Money supply Real GDP ATS billion Logarithmic scale (1830=100) % 100 1,000 30 25 20 15 10 10 5 0 –5 1 100 –10 1938 1940 1942 1944 1946 1948 1950 1952 1954 1938 1940 1942 1944 1946 1948 1950 1952 1954 Currency in circulation Index (left-hand scale) M1 Growth rate (right-hand scale)

Source: OeNB. For further information on the data, see annex.

The currency reform involved money supply to rise again. Ownership several steps: The Resumption of Pay- and use of foreign exchange remained ments Act of July 3, 1945, blocked 60% restricted. of bank deposits; withdrawals of the The lack of goods and food and the remaining 40% were allowed subject continued existence of a sizeable mone- to limits and for certain purposes only. tary overhang caused inflation to rise The Schilling Act of November 30, 1945, strongly and swiftly in the postwar reinstated the Austrian schilling as legal years. Inflation peaked at 97% in 1947. tender. In December 1945, reichsmark The Currency Protection Act of 1947 notes (and military schillings) in de- finally tackled the inflation problem. nominations of 10 reichsmark and above Its purpose was to reduce banknote cir- were exchanged into Austrian schillings culation and bank deposits, and to pre- at a rate of 1 to 1, up to 150 Austrian pare the cleaning up of bank balance schillings per person. Amounts beyond sheets. In December 1947, notes and this were transferred to limited access coins in circulation were exchanged accounts. However, the subsequent into new schilling notes at a ratio of release of blocked accounts as well as 3:1, except for an amount of 150 schil- the financing by the central bank of lings per person, for which a ratio of sizeable occupation costs caused the 1:1 was used. The accounts blocked by

MONETARY POLICY & THE ECONOMY Q3– Q4/16 17 A (not so brief) history of inflation in Austria

the Resumption of Payments and Schil- GDP growth to 0.1% in 1952, which ling Acts were completely rescinded, was, however, followed by a renewed and restricted accounts were converted vigorous recovery in subsequent years into claims against the federal govern- (4.4% in 1953 and double-digit growth ment. In effect, the measures halved rates in 1954 and 1955). The recovery the volume of both notes in circulation in 1953 was helped by repeated cuts in and of deposits. Furthermore, all in all central bank interest rates and by an five price-wage agreements among the effective depreciation of the Austrian social partners (i.e. employers’ and em- schilling in the context of exchange ployees’ associations) concluded be- rate unification, while the inflationary tween 1947 and 1951 aimed at contain- impact of exchange rate depreciation ing inflationary pressures.7 Inflation re- was contained by food price subsidies acted immediately and strongly, falling on wheat and imported animal feed; to 19% in 1948. The currency reform without these measures, inflation underpinned the Austrian schilling’s would, according to WIFO (1953), exchange rate, dampened black market have reached 3.7%. food prices and created a generally optimistic mood, in turn spurring a 5 Postwar recovery, post- strong economic recovery. Real GDP Bretton Woods era, hard currency policy and partici jumped by 28% in 1948, and posted re- pation in the euro area markable increases of 19% and 12%, respectively, also in the two subsequent The final institutional step to ensure years. By 1951, Austrian real GDP had lasting currency stabilization was taken doubled against 1946, clearly surpass- in 1955 with the passage of the National ing the prewar level. At the same time, bank Act, which contained a number inflation remained broadly under con- of provisions to increase central bank trol, averaging 20% in the four years up independence and which changed the to and including 1952. central bank’s mandate toward main- To counteract the chronically high taining both the domestic purchasing current account deficits, split exchange power and the international exchange rates were used after WW II until value of the Austrian schilling. While 1953, and the Austrian schilling was independent, the central bank was re- devalued in 1949 and 1950 ‒ however, quired to consider the economic policy without much effect. To curb the very of the federal government in exercising strong credit growth that had been ob- its monetary and credit policies (for de- served since 1947, credit controls were tails, see 1955 Nationalbank Act, Arti- adopted starting from 1951 ‒ unlike cle 2(2) and Article 4). the devaluations, these were effective.8 The success of currency stabiliza- Halting the credit expansion was the tion was confirmed by the moderate in- final step in breaking inflation: It fell to flation average of 2.3% in the years be- –0.7% in 1953. The decisive disinfla- tween 1954 and 1960, during which tion led to a temporary sharp fall in real GDP increased by 7% on average.

7 While the first three agreements were widely regarded as successful in containing inflationary pressures, the effects of the fourth agreement was questionable and the fifth agreement might even have contributed to higher inflation (Butschek, 2011). 8 The Korean War in the early 1950s spurred U.S. military expenditure and inflation, which had global repercussions. Part of the inflation spike in Austria in 1951 may also have been related to these developments (Breuss, 1980).

18 OESTERREICHISCHE NATIONALBANK A (not so brief) history of inflation in Austria

Chart 8 Post-war recovery, post-Bretton Woods, hard currency policy and euro adoption (1955 to present) Price index Inﬂation Index (1925=100, logarithmic scale) % Standard deviation in % 10,000 10 3.0

8 2.5 2.0 6 1,000 1.5 4 1.0

2 0.5

100 0 0.0 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 1955 1965 1975 1985 1995 2005 2015 Price index Inﬂation rate (left-hand scale) Inﬂation volatility (right-hand scale)

Nominal GDP and money supply Real GDP ATS billion EUR million Index (1830=100, logarithmic scale) % 10,000 1,000,000 10,000 15

10 1,000 100,000

1,000 5

100 10,000 0

10 1,000 100 –5 1955 1965 1975 1985 1995 2005 2015 1955 1965 1975 1985 1995 2005 2015 Currency in circulation (left-hand scale) Index (left-hand scale) M1 (left-hand scale) Growth rate (right-hand scale) Nominal GDP (right-hand scale)

Source: OeNB. For further information on the data, see annex.

The 1960s were characterized by tran- government and the social partners, sition to full employment and relatively which aimed at establishing a fine bal- pronounced fluctuations of prices and ance between international cost com- wages. Between 1961 and 1971, infla- petitiveness and real wage gains and at tion averaged 3.7%, and real GDP ensuring social well-being and peace. growth averaged 4.5%. Toward the end When inflation accelerated in 1957, of the 1960s, inflation accelerated in the Parity Commission for Wages and the wake of global inflationary pres- Prices was established to control and sures emanating from U.S. expendi- dampen wage and price developments. ture for the Vietnam War and as a re- Social partners committed themselves sult of high wage increases in 1968 and to reporting all planned price increases 1970 (Breuss, 1980). and wage demands to the commission The combination of a remarkable for evaluation (Breuss, 2011). In the be- economic catching-up process with rel- ginning of the 1960s (Raab-Olah-Ab- atively moderate inflation during the kommen) the responsibility for wage two decades until 1971 was facilitated negotiations was transferred to a sub- by the close cooperation between the committee on wages9 that approved the

9 A subcommittee on prices had already been established in the 1950s.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 19 A (not so brief) history of inflation in Austria

result of wage negotiations (system of of the U.S. dollar’s international con- collective wage bargaining in Austria).10 vertibility into gold in August 1971 In addition, investment tax rebates and marked the end of the postwar Bretton favorable investment financing helped Woods system of fixed exchange rate develop the capital stock and bolstered parities. Unlike Germany, which let potential growth. Export support the Deutsche mark float from May schemes, which included favorable pub- 1971, Austria feared that free floating lic guarantee schemes and financing, would expose the schilling exchange fostered Austria’s integration in the rate to international speculation. In- global economy. stead, all political forces in Austria de- Since the adoption of a unified ex- cided jointly to revalue the Austrian change rate in 1953, the Austrian schil- schilling by 5.5%. The revaluation was ling’s parity vis-à-vis the U.S. dollar aimed at curbing inflationary pressures and thus also gold was maintained (inflation stood at 4.4% in 1970 and within a margin of ±1%. This implied, 4.7% in 1971). When many European however, a devaluation of some 15% Community countries adopted a policy against the Deutsche mark – the cur- of block floating in 1973 and the Deut- rency of Austria’s main trading partner sche mark was revalued by 3%, Austria – and revaluations of 15% and 37%, re- opted for a 2.25% revaluation, while spectively, against the pound sterling retaining its orientation toward the and the French franc. The suspension basket of currencies of its main trading

Chart 9 Largest correlation between inﬂation and currency in circulation or M1 Correlation coeﬃcient 1.0 –2 –2 1 0.8 –2 2 –2 –2 0.6 –2 –2 –2 0.4 4

0.2

0.0

–0.2

–0.4 –5 0 –0.6 5

–0.8 4 –6

–1.0 1952 to 1971 1972 to 1998 1952 to 1971 1972 to 1998 1952 to 1971 1972 to 1998 1952 to 1971 1972 to 1998 Inflation and currency Inflation and currency Inflation and M1 Inflation and M1 (+lags/–leads) in circulation in circulation (+lags/–leads) No filter 2 to 8 years 8 to 20 years 20 to 40 years

Source: Authors’ calculations. Note: Figures attached to bars show lags(+)/leads(–) in years.

10 Even though the Parity Commission has only an informal basis, the Commission played an important role in Austrian economic policy. For an analysis of the functioning of the Parity Commission see Marin (1982).

20 OESTERREICHISCHE NATIONALBANK A (not so brief) history of inflation in Austria

partners, the so-called Indicator.11 A ing to 5.7%. This spike was triggered further revaluation by 4.8% was de- among other things by an increase in cided in June 1973 (again slightly less value added tax (Pollan, 1984). Infla- than the Deutsche mark’s revaluation tionary developments were also re- of 5.5%). In July 1976, the peg was flected in relatively high inflation vola- switched to the Deutsche mark, with tility. At the same time, 1975, 1978 some minor downward and upward ad- and 1981 were recession years in justments until 1981. Austria, with real GDP growth turning Regarding the empirical relationship slightly negative for the first time since between money growth and inflation, WW II. Economic theory only in hind- the two post-war decades up to the end sight characterized the joint appearance of the Bretton Woods system in Austria of high inflation and low or negative fitted very well the textbook notion of a economic growth as “stagflation.” At weak relationship in the short run but a the time, Austrian economic policy- very strong positive relationship in the makers, like those in most other coun- long run. During those two decades, tries, initially attempted to cushion the most correlation coefficients for me- negative growth impact of the oil price dium- and long-run frequencies were shocks by taking expansionary fiscal higher than those found for the whole policy measures (note that the authori- sample (annex) and for the other three ties’ choice to pursue an exchange rate subperiods for which we conduct this peg meant that monetary policy was analysis. What is more, we find that not used actively and autonomously). during this period, the correlation be- The result was the described rise in in- tween growth of monetary aggregates flation and a doubling of Austria’s pub- and inflation was positive when mone- lic debt ratio during the 1970s along- tary aggregates led inflation. Our em- side sagging GDP growth and rising pirical findings for this period are thus unemployment (albeit probably by less in line with the notion that inflation is a than without the fiscal measures, at monetary phenomenon (chart 9). least in the short run). Again, wage The demise of the Bretton Woods moderation was used actively to pre- system in 1971 and the first oil price vent a wage-price spiral from develop- shock in 1974 spurred inflation glob- ing as a result of the one-off price hike ally. Credit growth in Austria was in oil. buoyant, prompting the OeNB to take To fight rising inflation, exchange various restrictive monetary policy rate policy played an increasingly cen- measures and to deliberalize interna- tral role from the 1970s. The peg to the tional capital transactions. Inflation in Indicator and from 1976 to the Deut- Austria reached 6.4% in 1972 and sche mark implied that the Austrian peaked at 9.5% in 1974 before gradu- schilling’s nominal effective exchange ally receding to 3.6% by 1978. The rate gradually and consistently appreci- second oil shock triggered a renewed ated by a total of 80% between 1971 rise in inflation to 6.3% in 1980 and and 1995. Pegging proved effective in 6.8% in 1981. A final spike in inflation bringing inflation to below the levels of occurred in 1984, with inflation surg- Austria’s trading partners; as a result,

11 For details, see Handler (2016) and Schmitz (2016) in this volume.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 21 A (not so brief) history of inflation in Austria

the real effective exchange rate of the Three of EMU, which include low in- Austrian schilling appreciated much flation and sound public finances, be- less, but still substantially, climbing by came an economic policy priority. around 30% over the same period During the 17 years of Austria’s and exerting continuing pressure on euro area membership between 1999 Austria’s economy to improve produc- and 2015, inflation in Austria averaged tivity. During the period between 1983 1.9% and was thus in line with the ECB and 1993 – i.e. the period after the in- Governing Council’s definition of price flationary impact of the second oil price stability. Inflation during this period shock had been absorbed but before was also remarkably stable, exceeding Austria acceded to the European Union 3% only in two years (2008 and 2011) – inflation averaged 3.1% a year, i.e. on as a result of renewed oil price shocks. average 0.9 percentage points above the Around the time of the euro cash rate in the exchange rate-anchor coun- changeover (which took place in 2002) try Germany. The successful stabiliza- the perception of higher inflation was tion of the inflation rate was reflected relatively widespread. As the cash in a downward trend of inflation vola- changeover in fact did not affect mea- tility. EU accession brought inflation in sured inflation, perceived inflation and Austria down further to 1.9% on aver- measured inflation diverged widely age during 1995 to 1999, as the open- during that period. (Fluch and Rumler ing up of the Austrian economy in- (2005) and Fluch and Stix (2005)). In- creased competitive pressures in many flation was also rarely below 1%: in sectors, on the one hand, and as com- 1999 as a result of the Asian crisis and pliance with the Maastricht conver- the low in oil prices, in 2009 in re- gence criteria for participation in Stage sponse to the Great Recession, with

Chart 10 Phillips curve from 1964 to 2004 Regression coeﬃcient on output gap 1.2

1.0

0.8

0.6

0.4

0.2

0.0

–0.2

–0.4

–0.6

–0.8 1964 1969 1974 1979 1984 1989 1994 1999 2004 Coeﬃcient Coeﬃcient ±2 standard errors

Source: Author’ calculations based on OeNB data. For further information, see annex.

22 OESTERREICHISCHE NATIONALBANK A (not so brief) history of inflation in Austria

real GDP shrinking by 3.8%, and in in currency in circulation or M1 is 2015 as a result of the oil price col- more often negative depending on the lapse.12 frequency and the lags included, and Rolling regressions for the Second the high correlation between inflation Republic (after 1945) suggest a posi- and M1 growth found for the pre-Bret- tive short-term correlation between ton Woods era is no longer found in the the output gap and the inflation rate post-Bretton Woods/pre-EMU era. This (chart 10). Regression results for the may reflect the hard currency regime whole postwar period (table A1 in the and the resulting endogeneity of money annex) corroborate this result. How- supply as well as the generally observed ever, for the period from 1972 to 1998, gradual breakdown of the stable rela- i.e. the time of the Austrian hard cur- tionship between money growth and rency policy after the end of the Bret- inflation in the course of the quarter ton Woods system, the coefficient on century up to 1998. the output gap is not significantly different from zero. Furthermore, the 6 Summary and conclusions role of the output gap drops if changes In this paper, we distinguished five dis- in crude oil prices are also included tinct periods of inflation history for among the explanatory variables (table Austria: the “long 19th century” with a A1 in the annex). The crude oil price fairly stable price level, partly relatively represents an external cost shock, i.e. high short-term inflation volatility and imported inflation. Exogenous price the frequent occurrence of negative shocks and changes in commodity inflation rates, which were not neces- prices were mentioned by e.g. Breuss sarily associated with economic con- (1980)13 as one of the reasons why the tractions; the two world wars, during Phillips curve vanished. Breuss (1980) which the central bank was forced argues that after the first oil crisis in to finance war expenses, resulting in 1973/74, the impact of imported cost- ultra-high or even hyperinflation; the push shocks on domestic inflation be- interwar period, which was marked came more important.14 by restrictive macroeconomic policies Looking at monetary aggregates, under the umbrella of the League of the strong empirical relationship ob- Nations and the gold standard as well as served for the two pre-Bretton Woods by the Creditanstalt and other bank- decades does not hold during the ing crises, leading to the Great De- post-Bretton Woods era (chart 9). The pression; and finally the post-WW II correlation between inflation and growth episode, marked by Austria’s remark-

12 In 2015, the inflation rate in the euro area as a whole was zero and even reached negative territory in some months. To avoid risks from longer periods of low or even negative inflation at a time when monetary policy rates have reached their effective lower bound, the ECB implemented various new measures (unconventional monetary policy). Analyses of recent inflation developments in Austria are published in the OeNB quarterly „Inflation aktuell“ (www.oenb.at/Publikationen/Volkswirtschaft/inflation-aktuell.html). 13 Breuss (1980) is one of several analyses from the 1970s and 1980s that discussed the stability of the Austrian Phillips curve. In contrast to our analysis, Breuss (1980) investigated the relationship between the unemployment rate and inflation (for the years from 1955 to 1978) and observed that this relationship is more stable in Austria than in most other countries. 14 Including other variables leads to a collapse of the Phillips curve relationship, suggesting that our results for the time of the monarchy (in particular from 1830 to 1866) could also be altered by including further variables and that the observed correlation between inflation and the output gap might not be robust to such changes. However, we had no suitable data for this time period.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 23 A (not so brief) history of inflation in Austria

able two- decade long economic catch- erate short-term war financing in ing-up period upon completion of the the 19th century (with the exception currency reform and monetary stabili- of the Napoleonic wars) did not en- zation, Austria’s response to the demise danger monetary stability. The post- of the Bretton Woods system through WW I inflation may be squarely as- an evolving exchange rate peg, and, ul- cribed to the central bank’s contin- timately, Austria’s accession to the EU ued acceleration of the money print- and the introduction of the euro. De- ing press well after the end of the spite these quite distinct periods, some war for fear of political instability stylized patterns of inflation develop- and social unrest caused by monetary ments in Austria over the past two reform. In the end, delaying neces- centuries emerge from our analysis. sary monetary reform increased the • The large expenses implied by wars reform cost. The bulk of WW II in- were repeatedly used to justify mon- flation actually occurred only after etary financing by the central bank. the war, as price controls and goods Central bank statutes conferring management had tamped down infla- nominal independence from the state tion during wartime. But once infla- repeatedly turned out to grant insuf- tion soared after WW II, a series of ficient protection. Wars also implied mutually complementary measures sizeable negative shocks to productive were taken fairly swiftly to contain it. capacity. The combination of exces- • The absorption of the postwar mone sive money creation and reduced po- tary overhangs after the Napoleonic tential output repeatedly resulted in wars as well as after WW I and ultra-high or outright hyperinflation. WW II included three elements: • Rulers or governments were aware of first, conventional fiscal measures the ultimate consequences when they such as tax hikes or spending cuts; resorted to monetary financing. They second, ultra-high or hyperinflation accepted them deliberately, given that in particular wiped out the real their time preference for the short- value of private sector claims against term gains – mostly victory in war – the government; third, currency they hoped for. “reforms” that eliminated by law • The resulting unavoidable currency existing private claims against the reforms were politically hard and in- central bank (cash) or against banks cluded drastic measures with huge (bank accounts). The exact institu- distributive consequences that many tional form and the relative impor- citizens perceived as unfair and that tance of these measures varied over fundamentally undermined the social time (with the first measures being fabric. The financial losses that cur- least important), but the effect of all rency reforms inflicted on large parts measures was identical: They sharply of the population left people distrust- reduced existing private real finan- ful of state money and of state insti- cial wealth. The second and third tutions more generally for decades. types of measures fundamentally • Whether war financing led to ultra- violated one of the key functions of high or hyperinflation depended on money, namely of being a store of the volume and the length of the epi- value. Trust in state money therefore sode as well as on the speed and vigor suffered for long periods every time. of subsequent measures to contain • Disinflation as a rule also involved price pressures. In this vein, mod- real costs in the form of recessions

24 OESTERREICHISCHE NATIONALBANK A (not so brief) history of inflation in Austria

and higher unemployment; in this nized that stable money is beneficial sense, a Phillips curve-type relation- and that the central bank should be ship seems to have applied during such put in a position to ensure it. How- periods. At the same time, monetary ever, in the short term, immediate stabilization was the precondition for financing desires (mostly but not ex- the economic recovery thereafter. clusively for wars) often superseded • According to our empirical esti- the desire for monetary stability. mates, a Phillips curve relationship Austria’s monetary history shows between some measure for the out- that within a single state, statutory put gap and inflation in the short run independence does not provide ef- is statistically significant only during fective protection against monetary some periods. It seems to have been financing, excessive money creation weakest in Austria in the two post- and ultra-high or even hyperin Bretton Woods decades. flation, as the central bank statutes • Supply shocks dominated inflation can simply be changed at any time. developments in the post-Bretton By contrast, the experience of the Woods hard currency era. The high- Oesterreichisch-ungarische Bank sug- est nonwar-related inflation rates in gests that a “supranational” central Austria occurred in the 1970s as the bank within a monetary union of combined result of oil price cost-push several states may, despite all the shocks and the attempt to cushion complications involved in such an the negative permanent implications arrangement, de facto strengthen of the associated drop in potential central bank autonomy. output on economic growth and em- • The formal Maastricht rules for ployment through expansionary fis- sound fiscal policies and the strong cal policy. EU and EMU accession statutory independence of the Euro represented positive supply shocks. system go beyond any formal legal Also during Austria’s euro area mem- frameworks previously in existence bership positive and negative oil price in Austria in ensuring monetary sta- shocks were a major driving force be- bility. However, considering the poli- hind variations in the inflation rate. cies actually practiced since the in- • Very low or even negative inflation troduction of the hard currency was not systematically associated policy, the de facto regime change for with low real GDP growth in the 19th Austria resulting from EU and EMU century. By contrast, in the 1930s accession appears more modest: For during the Great Depression, defla- three decades, Austria had already tion was accompanied by high unem- gained experience in managing a ployment and economic slack. fixed exchange rate and employing • Financial bubbles also boosted con- other instruments of economic policy sumer price inflation; the ensuing to ensure international competitive- busts were accompanied by extended ness and to avoid unsustainable mac- periods of deflation. roeconomic imbalances. • Throughout the entire 200-year his- • In Austria’s monetary history, do- tory of Austria’s central bank, law- mestic consumer price stability in the makers and central bankers recog- sense of modern inflation targeting

MONETARY POLICY & THE ECONOMY Q3– Q4/16 25 A (not so brief) history of inflation in Austria

rarely played a role as an operational the period since Austria’s member- target.15 Most of the time, some form ship in the euro area. of external anchor – the price of sil- • Looking at the relationship between ver or gold during the 19th century money growth and inflation, the and the interwar period, a peg to frequency domain analysis conducted in another currency (or briefly a cur- this article for the pre-WW I and rency basket) in the post-WW II post-WW II periods finds that, in period –was used to guide monetary line with the economic literature, for policy. long and very long frequencies, the • While inflation was quite volatile in correlation between money growth the decade after the end of the Na- and inflation is considerably higher poleonic wars and also in the two than for short business cycle frequen- decades following the revolution cies. We also find that the correlation of 1848, it was very low in the half between money growth and inflation century after the Gründerzeit boom. has changed over time, reflecting In the post-WW II period, infla- changing monetary regimes over the last tion volatility rose markedly during two centuries in Austria and the world. the period of the Great Inflation • Austria’s 200 years of monetary his- between the late 1960s and early tory was also marked by the quest 1980s, in parallel with global devel- for maintaining or re-establishing opments after the demise of the Bret- the credibility of the central bank’s ton Woods system and the first and ability and commitment to maintain second oil shocks. In line with inter- price stability. The most recent mea- national developments, Austria also sures by the Eurosystem to credibly witnessed much lower inflation vol- prevent inflation from falling below atility during the Great Moderation the definition of price stability also fit since the 1990s. This includes also in this more general pattern.

References Angeloni, I., L. Aucremanne and M. Ciccarelli. 2006. Price setting and inflation persistence: Did EMU matter? ECB Working Paper Series 597. Antonowicz, W., E. Dutz, C. Köpf and B. Mussak. 2016. Memories of a central bank. Oesterreichische Nationalbank. Since 1816. Brandstätter. Vienna. Bachinger, K., F. Butschek, H. Matis and D. Stiefel. 2001. Abschied vom Schilling. Eine österreichische Wirtschaftsgeschichte. Graz. Bank of Greece, Bulgarian National Bank, National Bank of Romania and Oester reichische Nationalbank (eds.). 2014. South-Eastern European Monetary and Economic Statistics from the Nineteenth Century to World War II. Benati, L. 2005a. Long-run Evidence on Money Growth and Inflation. Bank of England Quarterly Bulletin. Autumn. 349–355. Benati, L. 2005b. The Inflation-Targeting Framework from an Historical Perspective. Quarterly Bulletin. Summer. Bank of England. 160–168. Benati, L. 2009. Long Run Evidence on Money Growth and Inflation. Working Paper Series 1027. European Central Bank.

15 In fact, methods to measure consumer prices were not even evolved enough to use consumer prices as operational targets. Most of the currently available 19th-century (and earlier) indices were not constructed until the 20th century. For a detailed account of the history of price measurement, see Fluch (2016) in this volume

26 OESTERREICHISCHE NATIONALBANK A (not so brief) history of inflation in Austria

Brandt, H. 1978. Der österreichische Neoabsolutismus: Staatsfinanzen und Politik 1848–1860. Göttingen. Breuss, F. 1980. Gibt es eine stabile Phillips-Kurve in Österreich? Monatsberichte 4/1980. WIFO – Austrian Institute of Economic Research. Bruggeman, A., G. Camba-Méndez, B. Fischer and J. Sousa. 2005. Structural filters for monetary analysis: The inflationary movements of money in the euro area. Working Paper Series 470. European Central Bank. Butschek, F. 1996. Statistische Reihen zur österreichischen Wirtschaftsgeschichte. Die öster reichische Wirtschaft seit der Industriellen Revolution. Vienna. Butschek, F. 2011. Österreichs Wirtschaftsgeschichte. Von der Antike bis zur Gegenwart. Wien. Fluch, M. 2016. The measurement of inflation in Austria: a historical overview. In this volume. Fluch, M. and F. Rumler. 2005. Price Developments in Austria after EU Accession and in Monetary Union. In: Monetary Policy & the Economy Q2/05. OeNB. 69–89. Fluch, M. and H. Stix. 2005. Perceived Inflation in Austria – Extent, Explanations, Effects. In: Monetary Policy & the Economy Q3/05. OeNB. 22–47. Handler, H. 2016. Two centuries of currency policy in Austria. In this volume. Jobst, C. and T. Scheiber. 2014. Austria-Hungary: from 1863 to 1914. In: Bank of Greece, Bulgarian National Bank, National Bank of Romania and Oesterreichische Nationalbank (eds.). South-Eastern European Monetary and Economic Statistics from the Nineteenth Century to World War II. Jobst, C. and H. Kernbauer. 2016. The Quest for Stable Money. Central Banking in Austria. 1816–2016. Campus. Frankfurt. New York. Jobst, C. and H. Stix. 2016. Florin, crown, schilling and euro: An overview of 200 years of cash in Austria. In this volume. Kausel, A. 1979. Österreichs Volkseinkommen 1830 bis 1913. In: Österreichisches Statistisches Zentralamt (ed.). Geschichte und Ergebnisse der zentralen amtlichen Statistik in Österreich 1829–1979. In: Beiträge zur österreichischen Statistik 550. Österreichisches Statistisches Zentralamt. Vienna. 689–720. Kernbauer, H. 1995. Österreichische Währungs-, Bank- und Budgetpolitik in der Zwischenk- riegszeit. In: Talos, E. (ed.). Handbuch des politischen Systems Österreichs. Erste Republik 1918– 1933. Vienna. Kernbauer, H. 2016. Cashless payments in Austria: the role of the central bank. In this volume. Komlos, J. 1986. Die Habsburgermonarchie als Zollunion. Die Wirtschaftsentwicklung Öster reich-Ungarns im 19. Jahrhundert. Vienna. Köhler-Töglhofer, W., D. Prammer and L. Reiss. 2016. The financial relations between the Nationalbank and the government. In this volume. Marin, B. 1982. Die Paritätische Kommission. Aufgeklärter Technokorporatismus in Österreich. Wien. Meller, B. and D. Nautz. 2012. Inflation persistence in the Euro area before and after the European Monetary Union. In. Economic Modelling. 29(4). 1170–1176. Mühlpeck V., R. Sandgruber and H. Woitek. 1979. Index der Verbraucherpreise 1800– 1914. Eine Rückberechnung für Wien und den Gebietsstand des heutigen Österreichs. In: Geschichte und Ergebnisse der zentralen amtlichen Statistik in Österreich 1829–1979. Beiträge zur österreichischen Statistik 550. Österreichisches Statistisches Zentralamt. Vienna. Nationalbank Act. 1955. Bundesgesetz vom 8. September 1955 zur Neuordnung der Rechtsverhältnisse der Oesterreichischen Nationalbank. Bundesgesetzblatt. Retrieved on July 12, 2016: www.ris.bka.gv.at/Dokumente/BgblPdf/1955_185_0/1955_185_0.pdf Pollan, W. 1984. Hohe Inflationsrate in Österreich im Vergleich zur BRD im Jahr 1984. In: WIFO-Monatsbericht 11–12/1984. 694–703.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 27 A (not so brief) history of inflation in Austria

Robalo-Marques, C. 2004. Inflation persistence: Facts or artefacts? ECB Working Paper 371. Schmitz, S. W. 2016. The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator. In this volume. Seidel, H. and H. Neiss. 1966. Die Steigerung der Verbraucherpreise von 1953–1966. In: WIFO-Monatsberichte 39(8). 302–310. Shelley, G. and F. Wallace. 2005. The relation between U.S. money growth and inflation: Evidence from a band-pass filter. In: Economics Bulletin 5. No. 8. 1−13. Van Walrè de Bordes, J. 1924. The Austrian Crown. Its Depreciation and Stabilization. London. WIFO – Austrian Institute of Economic Research. 1953. Zur Vereinheitlichung der Wechselkurse. In: WIFO-Monatsberichte 4/1953. Zipser, W. 1997. Auf der Suche nach Stabilität: Das Zentralbankgeldangebot der österreichischen Nationalbank 1923 bis 1937. Frankfurt am Main.

Annex Jobst and Kernbauer, 2016). The split Time period for statistical and of the post-WW II period into the pe- econometric analyses riods before and after 1971 corresponds We exclude the 1914 to 1952 period to the end of the Bretton Woods system from our statistical and econometric and the hard currency policy adopted in analyses because inflation developments Austria instead. were too unstable to allow for meaningful statistical inference. Inflation data in Phillips curve that period are affected by many fac- We regress the inflation rate on the tors: WW I, the interwar period with output gap and lagged inflation. We hyperinflation, WW II, the wartime use output because data for the time be- economy, scarcity as result of the war, fore WW I are not available for other and the questionable reliability of offi- measures (e.g. unemployment). For the cial inflation data, e.g. because of price time before WW I, we use data on in- controls and illicit trade. According to dustrial production (Komlos, 1986) Seidel and Neiss (1966), stabilization that are available from 1830 as a proxy was completed in 1953 and a “normal” for output. Another series for indus- development began. For reasons of con- trial production as well as GDP data sistency and comparability over time, are available from 1870 (on these series, in this article we prefer models that can see also Jobst and Scheiber, 2014). Even be estimated over the entire 200-year though there are differences between period. Clearly, the main constraint is these output data, the main conclusions data availability. We explicitly recog- with regard to the Phillips curve are nize that more sophisticated estimation not affected. For the time after WW II, approaches would in principle be desir- we use GDP data. We compute the out- able for a rigorous analysis but would put gap using a Hodrick-Prescott filter. depend on the availability of the neces- The rolling regressions shown in the sary data. charts use 21 observations (years); the The rationale for splitting the result is assigned to the midpoint of the pre-WW I century into two subperiods estimation period. The number of ob- (1830 to 1866 and 1867 to 1913) is the servations included in the regression is transition from fiscal to monetary dom- large enough for reasonable analysis but inance around the time of the Com- also allows us to take into account vari- promise of 1867 and (see section 1 and ations over time.

28 OESTERREICHISCHE NATIONALBANK A (not so brief) history of inflation in Austria

Our approach is subject to several Table A1 caveats, e.g. that the reliability of out- Phillips curve: dependent variable inflation put gap estimations is in general questionable. In addition, in the 19th cen- Before WW I tury, it is particularly difficult to disen- 1830–1913 1830–1866 1867–1913 tangle one-off price level changes and Output gap 0.297*** 0.453** 0.160 an inflationary stimulus because, as a Lagged inflation 0.199* 0.220 0.277** result of the typically large agricultural Constant 0.007* 0.011 0.003 sector, good harvests would be inter- After WW II preted as an apparent rise in potential 1957–2014 1972–1998 output and bad harvests as an apparent 16 Output gap 0.387** 0.240 0.250 –0.022 fall. Furthermore, results are not al- Lagged inflation 0.771*** 0.742*** 0.869*** 0.797*** ways stable with respect to the estima- Crude 0.011** 0.011** tion technique applied (e.g. ordinary Constant 0.008** –0.003 0.004 –0.005 least squares, OLS, versus least abso- Source: Authors’ calculations based on OeNB data. Crude oil price: IMF. lute deviation, as in Benati, 2005b). Fi- Note: ***, **, and * denote significance at the 0.01, 0.05 and 0.10 level, respectively. Output gap calculated nally, we apply only a very simple using HP filter, before WW I based on industrial production, after WW II based on GDP. model that does not, e.g., take into account inflation expectations but that we are able to use for the entire 200- We use two indicators as proxy year period. variables for money: currency in circu- Table A1 supplements the rolling lation, which is available from 1819, regressions displayed in the charts in and M1, which is available from 1867. the main text. In the table for the The reason we concentrate on narrow Second Republic, we begin only with money is purely practical; time series 1957 (as compared to 1954 in the roll- for M2 and M3 are available only from ing regressions), as the oil price series 1953 onward, so it would not be possi- (source: IMF) that we use to examine ble to carry out the analysis for the 19th the stability of our results are only century. Moreover, the reason for the available from 1957 onward. correlation between money growth and inflation is based on the fact that money Frequency domain analysis is used to carry out transactions. The To complete our historical analysis of kinds of deposits normally included in inflation, we analyze the long-run rela- M3 (and M2) were not used for trans- tionship between a measure of money actions in the 19th century; thus, it growth and inflation. In the long run, a would not make much sense to carry strong correlation between money out such a statistical analysis based on growth and inflation is usually found, M2 or M3. while in the short run, inflation is bet- Two of the time series that we are ter explained by real and external using are discontinuous because no data shocks. While we describe the relation- for currency in circulation and M1 are ship between these variables for each available for the time of the two World subperiod, the statistical analysis is Wars. Even though data for inflation based on the long-run characteristics of in this period are available, using these our entire sample. data for the analysis we are interested in

16 We thank the referee for this comment.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 29 A (not so brief) history of inflation in Austria

Chart A1 Growth rates and ﬁltered series of data used in the frequency domain analysis Growth rates of price index, currency in circulation and M1 % 35

–5

–25

–35 1801 1821 1841 1861 1881 1901 1921 1941 1961 1981 2001 Price index Currency in circulation M1

Filtered series at business cycle frequency (2 to 8 years) % 40

–10

–20

–30

–40 1801 1821 1841 1861 1881 1901 1921 1941 1961 1981 Inﬂation Currency in circulation M1 Source: Authors’ calculations.

would dominate the relationships and These peaks are due to the statistical thus distort the results for “normal” pe- reclassification of these variables upon riods. Furthermore, an extensive liter- EMU implementation. Starting in ature already analyzes the relationship 1999, the definition of M1 and cur- between money growth and hyperin- rency in circulation in Austria changed flation in those years;17 we summarize to “Austrian contribution” to currency the insights of this literature in the his- in circulation and to M1 of the euro torical narrative part of this article. area aggregates; thus, these have been Chart A1 (upper panel) shows the more or less artificial variables since annual growth rates of the price index, 1999. Hence, we chose to perform the currency in circulation and M1. Note analysis here with data until 1998 only. the large peaks in growth (positive and The frequency domain analysis was negative) of currency in circulation and done with the commonly used Chris- M1 with the start of EMU in 1999. tiano-Fitzgerald filter with the full

17 See e.g. Jobst and Kernbauer (2016), Bachinger et al. (2001) and van Walré de Bordes (1924).

30 OESTERREICHISCHE NATIONALBANK A (not so brief) history of inflation in Austria

Chart A1 continued Growth rates and ﬁltered series of data used in the frequency domain analysis Filtered series at long-run frequency (8 to 20 years) % 15

–5

–10

–15 1801 1821 1841 1861 1881 1901 1921 1941 1961 1981 Inﬂation Currency in circulation M1

Filtered series at very long-run frequency (20 to 40 years) % 10

–2

–4

–6

–8

–10 1801 1821 1841 1861 1881 1901 1921 1941 1961 1981 Inﬂation Currency in circulation M1 Source: Authors’ calculations. sample, which has the advantage of us- flation. Filtered series for inflation using all the data available. ing the entire sample show an extreme For the frequency filter, a fictitious value for the hyperinflation years even continuous series was built that skips for the frequency of 20 to 40 years. the years for which data are missing. Table A2 provides an overview of Given that those are the hyperinflation the definition of currency in circulation years, we discard these years also for in- and M1 as well as the data sources.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 31 A (not so brief) history of inflation in Austria

Chart A2 Correlation between inﬂation and currency in circulation or M1 Correlation coeﬃcient 1.0

0.8 –4 4 0.6

0.4 –6 2 2 –6 4 0.2

0.0

–0.2

–0.4 6 –0.6

–0.8

–1.0 1830 to 1998 1830 to 1998 1830 to 1998 1830 to 1998 Inflation and currency Inflation and currency Inflation and M1 Inflation and M1 in circulation in circulation (+lags/–leads) (+lags/–leads) No filter 2 to 8 years 8 to 20 years 20 to 40 years

Source: Authors’ calculations. Note: Figures attached to bars show lags(+)/leads(–) in years.

Table A2 Currency in circulation in Austria by periods

Date Definition Source

1819–1847 Banknotes, Vienna standard florins (Wiener Währung) and Jobst and Kernbauer (2016, chart 3.4) treasury bills, without deposits at the OeNB 1848–1862 Banknotes, Vienna standard florins (Wiener Währung) and Jobst and Kernbauer (2016, chart 4.1) treasury bills, without deposits at the OeNB 1863(1867)–1913 M1 is defined as the sum of currency in hands of the public Bank of Greece, Bulgarian National Bank, National Bank of and demand deposits; the latter are net of interbank deposits. Romania and Oesterreichische Nationalbank (2014) M3 is M1 plus time and savings deposits. 1923–1937 Estimates. Banking sectors are not comparable. Zipser (1997)

1947–1998 M1 = banknotes and coins in circulation excluding gold and Butschek (1996) silver coins, excluding cash in vaults, excluding OeNB deposits and overnight deposits 1946–1950: Sum of banknotes, coins in circulation, overnight deposits (including checking deposit estimates from December 1951). Until 1950 including cash in vaults of banks. 1999–2015 EUR million OeNB statistics, Austria’s contribution to M3 and its counterparts in the euro area

32 OESTERREICHISCHE NATIONALBANK The measurement of inflation in Austria: a historical overview

Calculation of the first price indices at the international level began as early as the start of Manfred Fluch1 the 18th century. In Austria, consumer price indices (CPIs) have been consistently available since 1800. Though simple, the indices followed today’s basic structure, and chaining the various generations of indices allows for conclusions to be drawn on inflation rate trends, adjustments for inflation and indexation for a period now spanning more than 200 years. This article covers a historical period extending from the first calculations of inflation to the complex criteria that determine inflation measurement in Austria today. A historical analysis of the trend in Austrian household expenditures as a key determinant of the basket-of-goods structure of price indices illustrates the long-term shift in spending from basic necessities to upscale goods and services. In 2016, Austrian households are spending only around 25 % of their total expenditure on basic necessities. Over the past 100 years, the number of items in the basket of goods and services has increased 40-fold (from 20 to approximately 800 items), and the number of price reports that are included in the monthly inflation calculations has risen from fewer than 100 to more than 40,000 in 2016. Another point of focus is the causes of measurement inaccuracies – which have always existed in the calculation of inflation – and the methods for correcting them. Austria’s accession to the EU and, subsequently, Economic and Monetary Union (EMU) with its supranational monetary policy instigated a process of restructuring and, more importantly, harmonizing inflation measurement. The Harmonised Index of Consumer Prices (HICP), which has been computed since 1996 to supplement the national indices, is now the primary indicator of inflation under the Eurosystem’s monetary policy strategy. Finally, this article concludes with a discussion of the measurement factors required to be used in the calculation of the Austrian CPI and HICP under EU law as well as the conceptual differences still existing between these two inflation parameters. JEL classification: B10, E31, N3 Keywords: History of inflation measurement, Austria

The rate of inflation is a defining factor The most comprehensive description of in the economic cycle and in economic inflation measurement is the Consumer policymaking. It serves as a central price index manual – Theory and practice indicator of monetary policy in its func- (CPI manual; ILO et al., 2013). Com- tion as a measure of monetary depreci- piled by various international organiza- ation, as a point of reference in wage tions and institutions, the approxi- setting, as an adjustment factor for a mately 550-page manual comprises an variety of fees and rates, as an indexing extremely in-depth analysis of all tool for both commercial and private aspects of the complex calculation of contracts and as a deflator for a number inflation rates as performed by national of price-adjusted calculations of eco- statistical offices. The guidance pro- nomic indicators (such as economic vided in this manual is based on knowl- growth). The measurement of inflation edge and experience reaching back to has therefore always required the high- the start of price measurement in the est possible degree of professionalism 18th century, which over the ensuing and precision. years and decades was gradually devel-

1 Oesterreichische Nationalbank, Agenda Office – Governing Board, General Council and General Meeting, Refereed by: [email protected]. The views expressed in this paper are exclusively those of the author and do not necessarily reflect those of the OeNB or the Eurosystem. The author would like to thank the referee, Josef Auer (Statistics Roman Sandgruber, Austria), Christian Beer, Clemens Jobst, Doris Prammer, Susanne Steinacher and Karin Wagner (all OeNB) for Johannes Kepler their helpful comments and valuable suggestions. University Linz

MONETARY POLICY & THE ECONOMY Q3– Q4/16 33 The measurement of inflation in Austria: a historical overview

oped further, improved, backed by sci- paper money Wiener Währung) were being entific methodology, subsequently har- issued, and Archduke Joseph II was monized and anchored in law at national elected King of the Romans in Frank- and EU level. furt. This was also the year in which This article, which has been kept the influx of precious metal imports technically simple, offers an overview from the New World led Italian econo- of the development of inflation mea- mist Gian Rinaldo Carli to begin his surement and the main instruments for deliberations on price fluctuations, measuring inflation in Austria over the resulting in his attempt to index the up- past approximately 300 years. Section ward movement of prices in 1500 and 1 covers the origins of inflation mea- 1750. To do so, he selected three prod- surement in an international context ucts that were both significant and typ- and introduces the different schools of ical of the Italian economy and way of thought relating to the index concepts. life at that time: wine, wheat and olive Section 2 deals with the start of infla- oil. He then determined their prices in tion rate calculation in Austria and pro- the two years to be compared and cal- vides an overview of the price indices culated the average price index of the created to date. Section 3 goes into the three goods. Aside from minor flaws shifts that have occurred in spending (the basket of goods was too small, and patterns in a comparison covering the goods were not weighted), the re- 150 years. Section 4 constitutes the sults obtained by Carli made him one of heart of this study and describes the the first scientists to have successfully major features of the price indices used indexed price trends. However, Diewert in Austria in the 20th and 21st century. (1988) mentions William Fleetwood, Section 5 references the Boskin Report the English Bishop of Ely, as being the published in the U.S.A. in the mid- first person to have computed a price 1990s and comprises an analysis of index with constant weighting, which sources of error and measurement dif- Fleetwood described back in 1707 in his ficulties in the calculation of inflation book, Chronicon preciosum. Fleetwood with regard to the Austrian CPI and compared the prices of items in a basket HICP. Section 6 focuses on the harmo- of goods (including wheat, beer and nization process in inflation measure- clothing) from 1707 with the prices ment in the context of Economic and from 1440. But the father of indexed Monetary Union (EMU) and the euro figures, according to Diewert (1988), is area and on the associated implications Joseph Lowe, a Scottish journalist and for the Austrian price indices. Section 7 political economist who made crucial summarizes the core messages of the advances in index theory at the start of study and offers a brief assessment of the 18th century. Nicolas Dutot (1738) the direction that the measurement of and William Stanley Jevons (1863) like- inflation is currently taking. wise developed index formulas that are still in use today. 1 Inflation measurement: origin Some 100 years after Carli (i.e. and schools of thought around 1875), Ernst Louis Étienne 1.1 G. R. Carli: one of the fathers of Laspeyres and Hermann Paasche, both inflation measurement German economists and statisticians, In 1764, the Seven Years’ War had just created indices with weights (consump- ended, the first Vienna Bancozettel tion) for calculating price changes in (which later became the government Germany. Later, so-called “superlative

34 OESTERREICHISCHE NATIONALBANK The measurement of inflation in Austria: a historical overview

indices” were calculated as well (an lated prices change, the COLI is less overview is available e.g. in the 2013 suited than other methods to providing CPI manual or from Hoffmann (1998)). information on inflation. In the case of the CPI (also referred 1.2 Consumer price index (CPI) and to as a “direct index” in the literature), cost-of-living index (COLI): the the quantity (expenditure weights) is CPI concept has prevailed in kept constant and only the change in practice price is observed, meaning that the CPI Economic theory has developed rather is not only more transparent, but also clear concepts of how the purchasing easier to interpret. CPIs – to stay with power of money should be measured. the officially known indices – are Indices are intended to reflect changes differentiated by whether they follow in the expenses incurred in maintain- the Paasche method or the Laspeyres ing a certain standard of living. From a method. Laspeyres compares a basket historical perspective, the measure- of goods in the base period with prices ment of inflation has followed two dis- in the reporting period and the base tinct schools of thought regarding index period, and Paasche takes the basket of theory: the cost-of-living index (COLI) goods in the reporting period and com- concept and the consumer price index pares it with the prices in the reporting (CPI) concept. period and the base period (ILO et al., The COLI embodies an economic 2013). The Paasche approach, however, concept that takes a benefit-based has the disadvantage that no or hardly approach. It assumes that consumers any historical price data are available will maximize benefits and minimize for many of the current goods that must costs when choosing between different be included in a present-day consump- combinations of goods and services. A tion model. Moreover, the necessity of COLI reflects the minimum costs nec- having to make backward projections essary to realize an equal benefit at of consumption patterns that are always prices po und p1 relative to the quantity kept up to date requires constant revi- consumed. Hence, the COLI measures sion of the index series, which is labori- the changes over time in the costs in- ous and difficult to interpret. However, curred by a representative household in the main reason that the Paasche index purchasing a bundle of goods without model is avoided is that it is hardly pos- sacrificing any benefits. The pool of sible to redesign the basket of goods representative goods and services is each year. The Laspeyres approach, on constantly being adapted to reflect cur- the other hand, has the disadvantage rent consumption patterns, and it also that maintaining a constant basket over allows for consideration of the effects of many years – as was the case until the substitution by households. However, end of the 1990s – does not reflect cur- the statistical implementation of this rent consumption patterns and can highly complex concept would involve therefore lead to an overestimation of having to collect data on consumer inflation. spending on a constant basis, which Economic science was long unable would require major resources. to provide a clear answer regarding the Under the COLI approach, the correct approach and the exact meth- bundle of goods changes in line with ods of collecting price data and deter- the changes in relative prices. Since mining changes in prices as well as both the bundle of goods and the re- regarding the frequency of adjusting

MONETARY POLICY & THE ECONOMY Q3– Q4/16 35 The measurement of inflation in Austria: a historical overview

the weights. Both COLIs and CPIs have wages in Austria since the 12th century. their strengths and weaknesses in mea- The first price indices calculated in suring inflation trends and both include Austria from the 1920s offer much more potential (mostly technical) bias factors recent information; subsequent back- that could effect a distortion in the in- ward calculations extended back to flation rate. In considering these advan- 1800. Historical trends in the various tages and disadvantages and – most sig- price indices are well documented in nificantly – the practicality of calculat- publications by Statistics Austria (and its ing and interpreting the inflation rate, predecessor institutes2).3 the “pure price concept” developed by Price indices are available from Sta- Laspeyres – now with annual updates tistics Austria for the entire period to consumption patterns – has pre- from 1800 to the present, and con- vailed, which ensures that only changes sumption patterns have been tracked in prices affect the rate of inflation. from the start of the 20th century to the Price indices are therefore defined present (table 1). Those two elements as follows: A price index is an indicator were combined to create a variety of representing the change in certain rep- price indices, from which the inflation resentative prices. It indicates how the rates for Austria were derived. The in- prices of the goods and services included dices were chained to provide continu- in a typical household’s consumption ous data up to the present.4 First, sec- basket have changed. The price index tion 3 delves into the changes in house- acts as a measure of the level of inflation hold spending patterns in Austria, and or deflation in an economy as a whole or section 4 then discusses selected aspects in parts of the economy. A reference of the respective indices. point (base 100) is defined for the index, and that reference point is used for 3 Household spending patterns: all subsequent indices. long-term shift from basic necessities to luxury items 2 Price series have existed in th To calculate price indices, it is neces- Austria since the 15 century, sary to collect reliable information on and price indices since 1800 household consumption patterns. The Among the many innovations that household consumption surveys carried occurred in the era of Holy Roman out periodically in Austria offer infor- Empress Maria Theresa was an official mation on household spending and also determination of prices in Austria, an serve to establish the contents and endeavor that some individual cities had structure of the baskets of goods used already undertaken earlier. Pibram to calculate the indices. An important (1938) prepared data on prices in Vienna source of historical data is the first con- and Klosterneuburg between 1470 and sumer survey carried out between Feb- the end of the 18th century. Emmerig ruary 1912 and March 1914 among (2015) offers an extensive overview of working class families in Vienna. The the literature describing prices and methods used hardly differed from pres-

2 Statistische Zentralkommission (Central Statistical Commission), Bundesamt für Statistik (Federal Statistical Office), Österreichisches Statistisches Zentralamt (Austrian Central Statistical Office). To avoid using varying designations, this article uses the present designation, i.e. Statistics Austria, except for bibliographical references. 3 Österreichisches Statistisches Zentralamt (1979a, 1979b, 1990). 4 Antonowicz et al. (2016, p. 88–89) for an illustrated representation of inflation rates since 1800.

36 OESTERREICHISCHE NATIONALBANK The measurement of inflation in Austria: a historical overview

Table 1 Price indices in Austria since 1800

Index / type of household Publisher Calculation period

Index of food costs necessary for Central Statistical Commission January 1921 to January 1925 four weeks in Vienna Index of increases in the necessary Central Statistical Commission January, March, June, October, total spending of a Vienna family December 1921 Index of the cost of living for one Joint Committee January 1922 to January 1926; person in Vienna based on weekly subsequent backward calculation to consumption 1914 by the Federal Statistical Office New index of changes in the cost Federal Statistical Office January 1926 to December 1938 of living in Vienna Consumer price index – backward Austrian Central Statistical Office 1800 to 1914 (annually only) calculation in 1979 Index of retail prices Austrian Central Statistical Office 1914 to 1959 (monthly) Cost-of-living index April 1938 = 100 WIFO (Austrian Institute of April 1946 to February 1959 for a four-member working class Economic Research) family in Vienna Cost-of-living index April 1945 = 100 CPI 1958 I – average working class Austrian Central Statistical Office to end-1965 household CPI 1958 II – four-member working class household CPI Austrian Central Statistical Office, every 10 years, or every 5 years 1966, 1976, 1986, 1996, 2000, 2005, from 2000 onward Statistics Austria from 2000 onward 2010, 2015 – average household HICP Statistics Austria every 5 years 1996, 2000, 2005, 2010, 2015 – average household

Source: Reports from the Central Statistical Commission (1921), reports from the Federal Statistical Office (1922), Statistische Nachrichten (1925, 1926, 1938), and various CPI revision brochures from Statistics Austria and its predecessor institutions (Österreichisches Statistisches Zentralamt, e.g. 1977, 1987, 1997; Statistics Austria, 2001, 2011). ent-day surveys as households had to able). The information presented in keep a record of accounts even back table 2 for the period from 1912 to 1955 then5. Each household received 20 was taken from various publications, crowns (the equivalent of approximately and the more recent information came EUR 105 in 2016) for participating, from the various CPI revision brochures making the survey considerably more (references see table 2). The data on well paid in relative terms than, for in- food expenses prior to 1912 was derived stance, the most recent such survey from the dual-volume publication (Aus- conducted in 2009/10, in which partic- trian Central Statistical Office, 1979a ipating households received coupons and 1979b) issued to commemorate the worth EUR 50 (Statistics Austria, 150th anniversary of the Austrian Cen- 2013). However, back then the data col- tral Statistical Office in 1979. lection period comprised two years A comparison of the data over the compared with just two weeks for the 150 years since 1869 shows a decrease 2009/10 survey and the 2014/15 survey in the originally high share of goods re- (for which the results are not yet avail- quired to meet basic necessities in favor

5 Today for purposes of household consumption surveys, Austrian households may choose between keeping traditional household accounts and online accounts (approximately 10% of households decide in favor of keeping online household accounts).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 37 The measurement of inflation in Austria: a historical overview

of more “elastic” items. Consumer dropped below 50% in 1935, and in spending thus follows Engel’s law6, the 1980s food accounted for approx- which states that as prosperity increases imately one-fourth of household through steady economic growth and spending. In the 2000s, the propor- rising per capita incomes, the proportion of food spending declined again tion of income spent on basic necessi- to around 12% (or approximately ties (such as food, shelter, lighting and 15% when including alcohol – i.e. heating) decreases, while that spent on beer and wine – and tobacco). goods with a high level of income elas- • Expenses for clothing previously ticity increases. This trend has been made up a greater proportion of particularly evident since 1955 with household budgets (between 10% “necessities” spending being replaced and 15%), whereas in 2016 only more and more by “luxuries” spending. around half as much (6%) is being However, the shift also reflects the spent on this item. transition from an industrial society to • Spending on education/culture and a service-based society in which spend- transportation (including tourism) ing on services is making up an increas- has risen steadily, increasing from ap- ingly greater portion of household bud- proximately 5% at the end of the gets. Basic necessities currently account 1960s to 13% (education/culture) for a share of only around 25% in and 17% (transport and tourism), re- household spending, i.e. the relation spectively, in the mid-1990s. Leisure between spending on basic necessities and tourism services have continued and nonessential items is 1:3. to increase sharply in significance up One major exception to this rule to the present day (accounting for is housing expenses, which since the approximately 26% of the HICP and 1960s7 have risen steadily in proportion 22% of the CPI in 2016). to household incomes. Limited real es- • In 2016, Austrian households are tate and increasingly higher standards of spending 53% of their budgets on housing quality as well as imbalances in goods and 47% on services. The pri- supply and demand and dynamic price mary items of expenditure are: food trends have all contributed to this effect. (11%) and “shelter, water, energy” The following additional trends in (CPI: 18%; HICP: approximately household spending patterns become 14%). Spending on energy alone ac- apparent in a long-term comparison counts for 8%. With respect to indi- (tables 2 and 3): vidual products and services, a high • In the late 19th and early 20th centu- proportion is spent on meat (3%), rent ries, Austrians spent around 60% of (5%) and vehicle purchases (4%). their income on food. That figure

6 In 1857, statistician Ernst Engel established that the percentage of income spent on food decreases as income increases. This effect, which has been confirmed for Austria as well, is known as Engel’s law (very generally, in economics the statistical relationship between changes in income and the resulting changes in spending in a specific necessities category is called the Engel curve). 7 The relatively low figure of 5% of total expenses for shelter in 1955 compared with the preceding decades could be related to the fact that the 1955 survey was the first one to compile spending data for all of Austria, whereas the previous data was restricted to Vienna.

38 OESTERREICHISCHE NATIONALBANK The measurement of inflation in Austria: a historical overview

Table 2 Change in consumer spending in Austria from 1869 to 1994

Consumption group 18691 18852 1912/14 19263 1935 1946 1955 1964 1974 1984 1994

% Food and beverages ⎫ 54.9 ⎫ 58 59.43 ⎫ 55.2 48.1 45.3 46.8 37.1 29.2 23.3 21.7 ⎬ ⎬ ⎬ Tobacco ⎭ ⎭ 1.5 ⎭ 1.9 8 7.6 2.0 2.6 2.5 1.7 Shelter 20.1 17 14.8 ⎫ 8.6 12.8 9.9 5.0 7.4 9.1 13.1 16.5 Lighting and heating 4 4.6 ⎪ 5.6 3.8 5.3 5.6 5.3 5.4 5.3 Furnishings and household ⎬ equipment 4.6 ⎪ 5.9 5.9 12.0 7.6 8.3 Clothing 12.5 15 9.3 ⎭ 22.3 9 5 13.1 12.7 12.9 10.7 8.7 Residential cleaning, etc. ⎫ ⎫ ⎫ ⎫ 2.2 2.7 1.9 1.4 1.5 Body and health care ⎪ 7 2.8 ⎪ 13.9 ⎪ 18.9 ⎪ 28.9 3.1 4.5 5.1 6.0 6.3 Education, leisure activities ⎬ ⎫ 7.7 ⎬ ⎬ ⎬ 6.8 8.3 9.2 14.3 13.4 ⎪ ⎬ ⎪ ⎪ ⎪ Transportation ⎭ ⎭ ⎭ ⎭ ⎭ 4.2 10.8 12.7 15.8 16.9

Source: Vienna Chamber of Labour (1955); Bundesamt für Statistik (1925, 1938); Österreichisches Statistisches Zentralamt (1977, 1979a, 1979b, 1987, 1997); for footnotes 1 and 2: Österreichisches Statistisches Zentralamt (1979b). 1 Four-member working class family in Vienna. 2 Five-member carpenter family in Vienna. 3 Consumer spending according to the 1926 COLI.

Table 3 Breakdown of consumption expenditure in the Austrian CPI and HICP from 2000 to 2016

COICOP1 consumption groups 2000 2005 2010 2015 2016

CPI HICP CPI HICP CPI HICP CPI HICP CPI HICP

% Total 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 Food and nonalcoholic beverages 13.61 13.11 12.23 13.04 12.01 12.75 11.78 11.67 11.70 11.34 Alcoholic beverages, tobacco 3.16 3.98 2.86 3.06 3.04 2.99 3.95 3.85 3.80 3.74 Clothing and footwear 6.97 8.27 5.46 6.35 6.14 5.64 5.91 7.19 5.70 7.15 Housing, water, electricity, gas and other fuels 17.92 13.93 17.92 14.58 18.65 14.55 18.48 14.18 18.40 14.13 Furnishings, household equipment and routine household maintenance 8.82 9.00 8.39 7.38 8.00 8.10 7.64 7.61 7.50 7.43 Health 3.51 1.91 4.64 4.07 4.44 5.38 4.81 5.08 4.90 5.21 Transport 14.26 14.52 15.13 15.14 13.78 14.16 13.33 14.50 13.20 14.42 Communication 3.09 3.21 2.66 2.65 1.81 2.13 2.22 2.19 2.20 2.12 Recreation and culture 11.99 11.63 12.42 11.24 12.13 11.37 11.43 11.18 11.40 10.99 Education 0.76 0.74 1.10 0.97 1.28 0.86 1.28 1.15 1.30 1.14 Restaurants and hotels 6.80 14.33 7.28 14.74 8.72 15.06 9.51 14.27 10.30 15.15 Miscellaneous goods and services 9.12 5.38 9.91 6.78 10.00 7.02 9.68 7.14 9.60 7.20

Source: Statistics Austria. 1 COICOP: Classification of Individual Consumption by Purpose. Note: Since 2000, the COICOP expenditure classifications have been used due to a change in the classification of consumption groups.

4 Price indices in the 19th, 20th kets of goods and the proportions of in- and 21st centuries: the basket of come spent on the various items in the goods grows, and survey and baskets. On the one hand, the number calculation methods improve of representative goods and services has The major expansion in both industrial been added to constantly, with a basket goods production and innovative spirit of goods containing a mere 20 items in is reflected doubly in the relevant bas- 1921, and 801 (CPI) or 789 (HICP) in

MONETARY POLICY & THE ECONOMY Q3– Q4/16 39 The measurement of inflation in Austria: a historical overview

2016 (tables 4 and 6), and on the other (Mühlpeck et al., 1979), for which the hand, the composition of the baskets survey among Vienna households of has changed dramatically, which has 1912–14 served as the basis for weight- also necessitated ongoing adjustments ing consumer spending. All of those in the individual household surveys indices evidenced deficiencies, however (replacing old items with new items, (e.g. an insufficient timeline, a limited particularly as a result of rapid changes basket of goods restricted to food, etc.), in quality criteria). This made it more meaning that a gap exists in the indices difficult, at times, to make consistent for the 19th century. comparisons of price indices and ensure To commemorate the 150th anniver- their continuity over several decades – sary of the Austrian Central Statistical a shortcoming that long-term index Office in 1979, a team of statisticians and inflation research still faces, albeit set about to calculate and document a in a less severe form. CPI index for the period from 1800 to This section outlines the various 1914. They gathered and organized the price indices in terms of selected fea- available data from the 19th century, tures of their index methodology and using price data for the cities of Vienna, changes in the structure of the relevant Graz, Linz and Innsbruck (including baskets of goods. The indices are listed data for different goods and consump- in chronological order (whereby the tion groups) for the entire backward first backward calculation to 1800 did calculation period. All in all, 13,000 not take place until the 1970s). prices for nearly 40 goods were processed. Missing prices were extrapo- 4.1 Backcasting consumer price lated or replaced with prices from indices from 1800 to 1914 nearby cities (as in the case of Linz with Price series for individual cities in the data from Wels). Divergent quantities various Austrian provinces have existed were standardized, and the prices were since the 14th century, and quite exten- converted, as a rule, to crowns and sive literature is available on the sub- hellers. The survey of 1912–14 was ject. At the beginning of the 18th cen- used here as a weighting model for the tury, even official price statistics were entire backward calculation period, still collected unsystematically. It was which appeared reasonable in light of not until the mid-19th century that in- the minimal changes in consumer terest in price series became greater in spending at that time (as documented connection with the 1873 World Fair in various 19th century sources). 1914 and the transition to a gold-backed cur- prices were set at 100 and taken as the rency in 1892. Initial attempts to calcu- index base. Price levels turned out to late aggregated price indices using price have increased from 62 to 100 over the data for a small number of products course of the 19th century, which cor- from the Habsburg monarchy’s markets responds to an inflation rate of only ap- also originated in that period. Not until proximately 1/2% per year. It must be during and after World War I (WW I), taken into account, however, that however, did statisticians begin con- prices were relatively high in 1800 structing backcasted indices of retail given that they originated in the time of and consumer prices in Austria the War of the Second Coalition be-

40 OESTERREICHISCHE NATIONALBANK The measurement of inflation in Austria: a historical overview

tween Austria and France, and 1914 parameters – an indication that the prices had not yet been impacted by the backcasted CPI is a sound indicator of impending WW I.8 inflation in the period, despite the res- With respect to the basket of goods ervations mentioned. compiled on the basis of consumer sur- Together with the retail price indi- veys of Vienna working class house- ces available from 1914 onward, a con- holds in 1912, the question arises as to tinuous price index – featuring chains whether that basket is representative of of subsequent generations of indices the entire period from 1800 to 1914, (which is no easy matter given the mul- especially with regard to the popula- tiple changes in currency) – is therefore tion outside of Vienna. On the one available for Austria. This made it pos- hand, statistics on food consumption sible to calculate fluctuations in mone- between 1780 and 1910 indicate little tary values for a continuous period change in per capita consumption pat- from 1800 onward as well as to resolve terns. However, by 1910 sugar was questions about indexing and adjust- already quite important (per capita ments for inflation in the 19th century. consumption of around 18 kg per year), whereas it was nearly unknown in 1800 4.2 Differing indices from 1921 to 1938 (annual consumption < 1 kg per capita). In 1921, the Central Statistical Com- In addition, rent expenses are also mittee for the first time recognized the likely to have accounted for a greater necessity of calculating an index in light proportion of spending in 1912 than in of the scarcity of food and rapid infla- 1800.9 The possible deficiencies in the tion prevalent in Austria at that time. index – a nonrepresentative basket of Due to the especially difficult living goods and lack of data on underlying circumstances and massive increases in regional consumption habits – led inflation, a number of different con- Cvrcek (2013) to calculate regional his- sumer price indices were created at torical cost-of-living indices for the ter- short intervals (table 1). The funda- ritory of the Habsburg monarchy from mental problem of the first official indi- 1827 to 1910 (on the basis of various ces was that apart from the basket of consumer goods available in that period goods to be determined for a house- and their prices). Cvrcek demonstrated hold, prices and price trends varied tre- that regional price levels differed mendously – in part due to state subsi- greatly, although they became more dies – and the Vienna prices were nor- similar over the course of the century, mally used in the calculations (Klezl, following similar trends and converg- 1925; Suppanz, 1976). Thus, the first ing near the end of the observation pe- indices were not at all representative of riod. When comparing the CPI with the rest of Austria. Not until 1958 did the COLIs used in Cvrcek’s study for the indices of consumer prices change (present-day) Austria, it ensures that to include regional data collected out- the rates of inflation show a similar side of Vienna. The basket used for the progression with respect to both start of index calculation in 1921 only

8 The index resulting from the backward calculation to 1800 does not consider the fact that hyperinflation in the Napoleonic era around 1800 had rendered paper money virtually worthless (Jobst and Kernbauer, 2016, p. 29, who refer to a 12-fold increase in prices at the beginning of the 19th century, which was not reflected in the index). The index taken from Mühlpeck et al. (1979) states all prices in either silver or crowns. 9 See Österreichisches Statistisches Zentralamt (1979a, p. 669; 1979b, p. 130–132).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 41 The measurement of inflation in Austria: a historical overview

included food and fuels, with expendi- was computed on a 1938 basis since it tures such as countryside sojourns, was still based on pre-WW II consump- baptisms and funerals added later. tion patterns. For the first time, the The first index in 1921 was estab- basket of goods included household lished to depict “food costs necessary items (silverware) and writing materi- for four weeks in Vienna.” The goal was als (ink, pencils, paper) as well as train to measure price movements at the fares and postal expenses. minimum subsistence level, i.e. food The COLI representing the ex- prices. The basket of goods represented penses of a four-member working class the subsistence necessary to supply a 70 kg family in Vienna (a price index as de- man with approximately 3,000 calories scribed in section 1.2; see WIFO, consisting of 70 g of protein per day. It 1949) was the direct predecessor of the varied from month to month depending later CPIs. WIFO calculated the COLI on the availability of goods for the pop- between 1946 and February 1959 using ulation. The index used the official two different base periods.11 The basket prices for rationed food quantities and of goods underlying the index was like- the prices on the free market and the wise based on surveys of the household black market for other goods. budgets of working class Vienna fami- In March 1921 the Central Statisti- lies from before WW II (specifically, cal Commission calculated a new index from 1935) and was broken down into of increases of the necessary total nine primary groups. Especially in the spending of a Viennese family. This immediate post-war period, the WIFO index was published for several months was skeptical about the index being a in 1921 (table 1). suitable representation of costs given In January 1922, the Federal Statis- that it initially underestimated the tical Office created an “index of the change in the cost of living, which at cost of living for one person in Vienna that time was very difficult to quantify, based on weekly consumption,” which and overestimated the price increases was calculated until December 1925. that occurred after currency reform in Starting in 1926 – once the economic 1948 and the wage-price-agreements situation in Austria had largely normal- between 1947 and 1951 (WIFO, 1949). ized again – that index was replaced with a COLI that remained in use until 4.4 CPI I 58 and CPI II 58 – price 1938. data also collected from the provincial capitals 4.3 RPI 38, COLI 38 and COLI 45 – Starting in March 1959, the COLIs inflation parameters for the were replaced by the new CPI series post-WW II period and index calculation was well docu- The 1938 retail price index (RPI)10 was mented from that time onward. The calculated in the post-World War II 1954/55 consumer survey provided (WW II) period between July 1948 information on the distribution of and February 1959 after the supply of spending for both indices. The index goods had largely normalized again and itself was calculated on the basis of state food rationing and price fixing 1958 prices using two different sets of had been lifted. However, the index data – one for an average household

10 March 1938 = 100. 11 COLI 38: April 1938 = 100; COLI 45: April 1945 = 100.

42 OESTERREICHISCHE NATIONALBANK The measurement of inflation in Austria: a historical overview

Table 4 Number of indexed items in the Austrian CPI from 1800 to 1986

Consumption groups 1800– 19212 19213 19224 19255 1938 1945 1958 1966 1976 1986 19141

RPI COLI CPI CPI CPI CPI

Total 38 20 40 24 44 54 83 197 253 582 615 Food and beverages 24 18 16 16 25 27 31 64 76 147 149 Tobacco 2 0 0 0 6 4 2 7 7 85 79 Construction, rent and maintenance of residential housing 1 0 1 1 1 0 1 3 5 25 28 Lighting and heating 5 2 5 4 4 4 6 10 10 11 13 Furnishings and household equipment 0 0 0 0 0 6 10 29 36 43 41 Clothing and personal care 3 0 10 3 3 3 16 27 35 44 50 Residential cleaning, laundry and dry cleaning 0 0 2 0 0 0 1 12 12 11 9 Body and health care 1 0 4 0 2 1 2 16 23 24 31 Leisure and education 1 0 1 0 2 6 10 19 30 71 85 Transport 1 0 1 0 1 3 4 10 19 121 130

Source: Statistics Austria. Österreichisches Statistisches Zentralamt (1977, 1979b, 1987, 1997). 1 Backward calculation from 1979 using data from the 1912–14 consumer survey. 2 Index of food costs necessary for four weeks in Vienna. 3 1921: Index of increases in the necessary total spending for a Vienna family. 4 1922: Index of the cost of living for one person in Vienna based on weekly consumption. 5 New index of changes in the cost of living in Vienna.

(CPI I) and one for a four-member dex that was representative of all house- working class family (CPI II). New to hold sizes, all income levels and all so- this index was, however, that in addition cial groups existing in the medium- to to Vienna and the Austrian provincial large-sized municipalities in Austria. A capitals at the time, price data were ex- total of 253 goods and services typical tended to include Wiener Neustadt and of the period were tracked each month. St. Pölten, which remained the case un- These included some rather innovative til 1976. Starting in 1958, the execu- products such as the portable type- tion of price surveys was based on an writer – the precursor of today’s lap- agreement between the Austrian Cen- tops and notebooks – the Puch moped tral Statistical Office and the partici- and the legendary VW Beetle (now re- pating municipal governments, with placed by the VW Golf and the VW the reporting cities receiving compen- Polo or revived in a much improved sation for their assistance. Retail estab- technical and qualitative form) as well lishments took part on a voluntary as eau de Cologne. With regard to ciga- basis. From 1966 onward, consumer rette brands, the Austria 3, Austria C price statistics were generally governed and Smart Export brands were at their by the Federal Statistics Act of 1965 peak and could be purchased for a price (Federal Law Gazette No. 91/1965). of between 25 and 30 Austrian groschen per piece. 4.5 CPI 66 to 96 – three-fold Consumer surveys were carried increase in the number of items out in ten-year intervals after the 1964 in the basket of goods survey, which was followed by the The consumer survey of 1964 provided 1974 and 1984 surveys from which the the foundation for the first general in- 1976 and 1986 generation of consumer

MONETARY POLICY & THE ECONOMY Q3– Q4/16 43 The measurement of inflation in Austria: a historical overview

Box 1

Comparison of CPI II, COLI and RPI from 1938 to 1958 The analysis of post-war price trends is based on the RPI 38 (1938 = 100) and COLIs 38 and 45 (1938 resp. 1945 = 100). The inflation trends depicted by these three indices can be compared using the CPI II 58, which like the COLI was based on a four-member Vienna family in terms of household size and had been calculated backward by the Austrian Central Statistical Office using a Paasche index. The RPI restricted spending patterns to those of an average adult in a low-income household. Noteworthy in this context is that the RPI and COLI series were based on extrapolations of the cost portions (quantity times price) of 1935 consumption patterns, but the CPI II was based on the 1954/55 consumer survey (fixed consumption patterns). In light of that difference and assuming a uniform base of 1954 = 100, we derive the trends shown in the table below. All three indices more than quadrupled between 1938 and 1951. While the CPI II with its varying index formulas only rose 4.4-fold, the RPI and COLI (Laspeyres indices) increased 4.9-fold.

Period CPI II1 COLI2 RPI2

1954 = 100 April 1938 16.2 14.7 14.7 1950 71.4 68.2 67.8 1951 88.1 87.1 86.5 1952 100.6 101.9 98.2 1953 97.2 96.5 97.5 1954 100.0 100.0 100.0 1955 101.4 100.8 102.5 1956 104.8 104.3 105.4 1957 107.2 106.6 109.7 1958 108.1 109.0 110.9 Change in % 1938–51 443.8 492.5 488.4 1951–54 13.5 14.8 15.6 1954–58 8.1 9.0 10.9

Source: Austrian Central Statistical Office, 1959. 1 1938–54 according to Paasche, from 1954 onward according to Laspeyres. 2 1938–58 according to Laspeyres.

price indices emerged. From 1976 on- (tables 2 and 4). This trend continued ward, all cities with more than 20,000 unabated in the decade between 1986 inhabitants (20 municipalities sur- and 1996 and resulted in another veyed) were included (the number of increase in the number of products retail establishments reporting grew relevant to spending patterns and in- from 1,350 to approximately 3,500; flation measurement. see table 5). Spending patterns were Starting in 1997, the Austrian CPI further refined, and approximately was converted to a 1996 = 100 basis. 600 goods and services were repre- The weighting was based on the sented in the CPI 76 – twice as many 1993/94 consumer survey, in which as ten years prior. The subsequent CPI around 6,600 households participated 86 did not contain significantly more by recording all of their household index items (615), despite the fact that spending for an entire month. The its composition indicated a further weighting structure of the CPI 96 shift in spending toward services was based on monthly expenditures of

44 OESTERREICHISCHE NATIONALBANK The measurement of inflation in Austria: a historical overview

Table 5 Selected figures on the Austrian CPI index elements from 1959 to 2010

1959 1966 1976 1986 1996 2000 2005 2010

Consumer survey Number Participating households 3,989 7,304 6,674 6,599 6.604 7,098 8,400 6,500 EUR

Monthly index-relevant expenditures per household 166 325 711 1,482 2,351 2,376 2,676 2,873

Index structure Reporting units Retail establishments n. a. 1,350 3,500 3,600 4,200 3,500 3,700 3,700 Price reports per month n. a. n. a. 30,000 33,000 40,000 39,000 41,400 41,400 Number

Number of municipalities from which regional prices were collected 10 10 20 20 20 20 20 20

Source: Statistics Austria. Note: n. a. = not available.

32,300 Austrian schillings; (around vidual Consumption by Purpose EUR 2,400) – the budget for an aver- (COICOP) structure already used for age 2.9-person household (or 1.8 in- the national accounts. Under the new come recipients in statistical terms). nomenclature, household consumption The basket used for the CPI 96 was broken down into 12 primary comprised 710 goods and services. Ap- groups, 40 groups of goods, and 106 proximately 40 items had been re- categories. The categories are broken moved and 165 were added. The num- down further by the individual EU ber of items under monthly observation Member States, depending on their increased to around 40,000 in some consumption habits, and represent the 4,200 retail establishments (table 5). level of the indexed items. An interim index (the predecessor of the HICP) in line with EU standards 4.6 CPI 2000 and HICP 2000 – was calculated for the first time in Jan- divergent design concepts uary 1996 (section 6). Starting in 1997, At the beginning of 2001, the index an improved version of the HICP was series was adjusted for the sixth time calculated by Statistics Austria in paral- since 1945 to reflect current spending lel to the CPI. Thus from 1996 onward, patterns and reset to a 2000 = 100 ba- two official inflation rates were avail- sis. The transition to five-year intervals able for Austria. However, the HICP instead of the earlier decade-long jumps found little public use due, in part, to was based on international recommen- the publication policy of Statistics dations, due above all to the high mo- Austria, which tended to favor the de- mentum and constant innovation exist- tailed results of the CPI in its monthly ing in today’s fast-paced product mar- press releases. In addition, the HICP kets. More than 7,000 households incorporated new spending categories, provided information on their monthly specifically the Classification of Indi- spending behavior in different months

MONETARY POLICY & THE ECONOMY Q3– Q4/16 45 The measurement of inflation in Austria: a historical overview

Table 6 Indexed items in the Austrian CPI and the HICP from 1996 to 2016

COICOP consumption groups in the 1996 2000 2005 2010 2016 CPI and the HICP

CPI HICP CPI CPI CPI CPI HICP

Number Total 710 678 812 770 791 801 789 Food and nonalcoholic beverages 125 125 127 129 128 128 128 Alcoholic beverages, tobacco 75 75 78 10 10 10 10 Clothing and footware 47 47 57 66 62 62 62 Housing, water, electricity, gas and other fuels 44 44 50 49 48 48 47 Furnishings, household equipment and routine household maintenance 64 64 74 75 72 72 72 Health 19 8 30 33 36 37 37 Transport 129 122 144 141 166 175 170 Communication 18 18 13 12 11 11 11 Recreation and culture 96 93 113 117 113 113 109 Education 7 4 12 13 15 15 15 Restaurants and hotels 32 31 40 47 51 51 50 Miscellaneous goods and services 54 47 74 78 79 79 78

Source: Statistics Austria. Note: Since 1996, the COICOP expenditure classifications have been used due a change in the classification of consumption groups. HICP index items are reported for the year of introduction (1996) and for 2016.

over one year. This survey and other presented a detailed comparison of the sources (box 2) were used to filter out two concepts as well as the index calcu- more than 800 typical household con- lations. The CPI and the HICP differ in sumer products to be used as the basis terms of the following criteria: for the CPI and the HICP. Additions 1. Consumer spending: residents con- were necessary in the area of transpor- cept (CPI) versus domestic concept tation and leisure activities in particu- (HICP)12, with the major difference lar, in which an especially wide range being that the HICP also accounts of new offers were available. For the for spending by foreign tourists in first time, the basket of goods and ser- Austria; vices included Internet fees, DVD play- 2. Items included in the basket of ers, CD-ROMs and computer printers goods and services: pursuant to EU as well as domestic help, child daycare regulations, the HICP does not in- and attorneys’ fees. clude owner-occupied housing, The revision of the consumer price gambling or parafiscal levies, for index and its conversion to a 2000 = which reason the two indices have a 100 basis in 2001 had an equal impact on different number of items and di- the CPI and the HICP due to the numer- vergent weights for consumption ous new regulations at EU level. Statis- groups; tics Austria therefore undertook exten- 3. Treatment of insurance policies: sive documentation of the differences gross concept (CPI) versus net con- between the two inflation parameters cept (HICP), i.e. the latter index for Austria (Statistics Austria, 2001) and takes insurance payouts to house-

12 Residents concept: consumer spending by residents of Austria; domestic concept: all consumer spending in Austria.

46 OESTERREICHISCHE NATIONALBANK The measurement of inflation in Austria: a historical overview

holds into account, which leads to a 4.7 The 2015 CPI and HICP: current much lower net expenditure weight inflation parameters for the HICP than the CPI; At the start of 2016, the CPI and the 4. Averaging methodology: arithmetic HICP were converted to a 2015 = 100 mean (CPI) versus geometric mean basis.14 The following section therefore (HICP) used for the index num- deals with measurement criteria that bers13 at the level of the elementary have not yet been mentioned in this index (e.g. long-grain rice in Linz); study and which are regarded as state of 5. Weighting adjustment: every five the art at present. years for the CPI versus annual The provisions relating to consumer weight checks and adjustments in price statistics (CPIs) are anchored in the form of a chain index for the the 200015 Federal Statistics Act and HICP (see below for details). other special regulations16. The HICP is The Laspeyres formula, the survey governed by a number of EU regula- methods (including the cities and the tions (section 6), with Regulation EC retail establishments included in the No 2494/95 supplying the framework price surveys), and the quality adjust- until end-2016. In May 2016, Regula- ment methods (section 5) remained the tion (EU) 2016/792 of the European same for the two indices. The geomet- Parliament and of the Council of 11 May ric averaging procedure (2005) and the 2016 on harmonised indices of consumer chain index (2010) used in the HICP prices and the house price index, and were later also incorporated into the repealing Council Regulation (EC) No national CPI. 2494/95 was adopted, repealing the The 2005 CPI and HICP revisions, previous regulation with effect of Janu- whose expenditure weights were based ary 1, 2017. on the 2004/05 consumer surveys, The Austrian CPI (and the HICP) is added brokerage account fees and flat- currently based on approximately screened TVs to the basket of goods 40,000 individual prices from 3,700 and services, and the previous 56 vari- retail establishments for around 800 eties of cigarettes were combined into products and services each month. By two items (cigarettes and cigars; see ta- comparison, approximately 1.5 million ble 6). Tobacco spending no longer individual prices are compiled per reached the threshold relevant for the month for the aggregate euro area CPI and was therefore not represented HICP. in the index.

13 The averaging methodology concerns the question of how prices are determined at the lowest level, i.e. the elementary index. In Austria, the prices for a certain good or index position (e.g. long-grain rice in Linz) in a certain city make up the regional elementary aggregate level. Around 16,000 elementary indices exist with respect to the 800 indexed items and 20 cities covered. Austria has used the index number method (arithmetic mean of the individual index numbers) since 1976 (before which the average price method was used). According to empirical studies carried out prior to designing the HICP, the HICP utilizes the geometric mean of the individual index numbers, which is better suited to depicting price trends. 14 Aside from the expenditure weights and the index items, it has not yet been possible to integrate any additional data or information from the 2014/15 CPI and HICP revision and conversion to a 2015 = 100 basis. The data from the 2014/15 consumer survey will not be available until the fall of 2016. 15 Most recently amended in Federal Law Gazette No. 40/2014. 16 Federal Law Gazette II No. 351/2003. Most recently amended in Federal Law Gazette No. 457/2015.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 47 The measurement of inflation in Austria: a historical overview

4.7.1 Not all prices are alike (b) are available to all potential con- Price data are generally collected in the sumers with no special conditions attached first week of the month, or in the third (non-discriminatory); week of the month for selected goods (c) are known to the purchaser at the (energy, seasonal merchandise). Quar- time when they enter into the agreement terly microcensus data are taken for with the seller to purchase the product rents and the percentage change in rent concerned; and prices indicated by the microcensus is (d) can be claimed at the time of pur- incorporated into the CPI. The prices chase or within such a time period follow- collected are the prices paid by house- ing the actual purchase that they might be holds for the acquisition (monetary expected to have a significant influence on transaction = purchase) of individual the quantities purchasers are willing to goods and services. The monetary purchase.” transaction is the deciding factor, i.e. Item (b) in particular has become the process by which a good or service more important in recent years given changes possession in return for money. the proliferation of retail customer loy- The actual purchase price includes all alty campaigns (Fluch et al., 2010). taxes applicable to the product and Such campaigns are excluded from cur- takes any rebates, discounts and special rent price measurement regulations be- benefits into account insofar as they are cause they are subject to the condition received by the consumer. Taxes in- that a customer card be issued and be- clude transportation taxes and excise cause the conditions are applicable to duties (e.g. motor vehicle, insurance individuals only. One particular issue and fuel taxes) and value added tax. in trying to include customer loyalty The HICP excludes parafiscal fees campaigns lies in the difficulty of iden- (while the CPI includes the en- tifying the campaign price and quanti- gine-based car insurance tax). Subsi- fying its relative importance for the dies, refunds and state grants are con- market as a whole. If, in an extreme sidered “negative” taxes and are de- case, a good or service were to be pur- ducted from the price. Transfers, gifts chased exclusively at the reduced price and savings are not taken into consider- based on all consumers possessing a ation because they do not involve mon- customer card, then that price would etary transactions. The prices are also be the representative price and should not impacted by the type or timing of be included in the index. The difficulty payment. in quantification lies in determining EU Regulation 2602/2000 governs which price is paid most often for a cer- the consideration of price reductions in tain product. The use of scanner data17 HICP calculation: in index calculation, which is currently “Unless otherwise stated purchaser in preparation in Austria, could effect a prices used in the HICP shall in general change in the collection of price data take account of reductions in prices of indi- since scanner data reflect the prices ac- vidual goods and services if such reductions: tually paid, including prices that are re- (a) can be attributed to the purchase of duced due to customer loyalty programs. an individual good or service; Certain sectors pose a particular challenge for inflation measurement;

17 Scanner data are already included in the current HICP calculations in other countries such as Belgium, Denmark, the Netherlands, Norway and Switzerland.

48 OESTERREICHISCHE NATIONALBANK The measurement of inflation in Austria: a historical overview

these include tourism (e.g. airplane an ongoing basis. A so-called “chain in- travel), residential rents, the communi- dex” makes this possible for the HICP cations sector (e.g. mobile phone fees) and, since 2010, for the CPI as well. and the educational and health care sec- The weights are adjusted annually and a tors as well as social institutions (e.g. standard procedure (weighting update) retirement home fees, etc.). These sec- is applied to update prices to reflect the tors demand particularly close atten- level recorded in December of the pre- tion due to their wide range of pricing ceding year18. Every five years, con- modalities and the difficulty of identi- sumer surveys and other sources of data fying quality changes where rates (box 2) are used to perform an exten- change quickly (box 4). sive revision of the goods and services included in the basket, with around 4.7.2 Criteria for retail establishments one-fourth of all index items being af- The following criteria apply to the re- fected by the changes. The 2010 revi- tail establishments from which price sion, for instance, involved adding 80 data are obtained: they must be repre- new goods and services, leaving out 59, sentative of as well as relevant to their and substantially adjusting the descrip- sector (i.e. have a certain market share) tion of 50 (table 7). Standard filter cof- and be regionally significant. All store fee makers were replaced by pod coffee categories must be included (small, makers in 2010, for example. GPS navi- large, specialty shops, supermarkets, gation devices, wood pellets and admis- discounters, catalogs, online shops, etc.). sion to thermal baths were also newly Approximately 90% of the price added. By contrast, unleaded regular data for CPI and HICP calculation are gasoline, color film and wall-to-wall collected at the level of the 20 Austrian carpeting were eliminated since those regions and 10% are collected directly products fell below the necessary thresh- from Statistics Austria (nationwide old of 0.1% of household spending. rates and fees, prices for especially The inflation rate calculated for complex products such as electronic Austria (no regional indices are pub- equipment). lished) is available around two weeks after the end of the reporting period 4.7.3 Basket of goods and services and is published by Statistics Austria. updated annually The inflation rates for all EU Member Short product cycles, constant product States are published by Eurostat on the innovation, the tendency of consumers same day. Approximately 120 indices of to adapt rapidly to new products and goods categories (plus special groups) take a critical view of “outdated” bas- are calculated in addition to the overall kets – all of which lead to distorted in- CPI and HICP. Statistics Austria addi- flation rates and do not adequately doc- tionally calculates and publishes special ument current consumer spending pat- price indices (including a micro and a terns – make it necessary to update mini basket, an energy price index and consumer spending data in the CPI on a passenger vehicle index)19.

18 The weights are adapted annually, i.e. the monthly indices for a given year are based on the previous year’s weighting structure, which is updated to reflect December levels and is “chained” to previous months via the month of December. 19 These indices aggregate index items from the various consumption groups: the passenger car index, for instance, tracks price trends such as automobile, tire and gasoline purchases as well as car repairs and insurance, etc. Until end-2015, a separate price index for retired households was calculated, but has been discontinued since.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 49 The measurement of inflation in Austria: a historical overview

Table 7 Annual revision of the basket of goods and services for Austria: from the 2005 CPI to the 2010 CPI

Basket of goods New items (80) Removed items (59) Substitute items (50) Basket of goods and services and services 2005 (770) 2010 (791)

Criteria: Criteria: Criteria: • Increasing sales • Sales decline • Increasing sales significance • Insufficient demand significance • New or additional • Same purpose purpose Examples: • >0.1% of consumer • Blast-furnace coke Examples: spending • Wall-to-wall carpeting • Pork ham instead of • Coffee filters processed sandwich Examples: • Unleaded regular meat • Pork loin gasoline • Rum instead of • GPS navigation • Color film brandy devices • Permanent waves • Pod coffee machines • Wood pellets instead of filter coffee • Gardening services machines • Coffee shop • Wheeled suitcases breakfasts instead of hard-shell • University courses suitcases • 24-hour care

Source: Statistics Austria.

Box 2

Derivation of expenses for the CPI and HICP baskets of goods and services: Consumer surveys and other sources Since the 1960s, the index baskets of goods and services have depicted the consumption of all households based on an average Austrian household. Since every household has its own individual consumption patterns, it would of course be quite unlikely in the real world to encounter a household precisely like the one underlying the calculation of inflation. For example, an individual household will not use all types of energy at the same time, but will have expenses for either gas, district heating or wood in addition to electricity. By the same token, few households will have expenses for both a rental home and an owner-occupied home. The index nonetheless includes those expenses on a proportionate basis. Comprehensive consumer surveys form the basis for the detailed breakdown of household spending but do not always supply reliable data, however. The data could be distorted due to recordkeeping errors, changes in spending behavior as a result of recording expenses, and the “whitewashing” of figures (data are not entered correctly in the household accounts – e.g. expenses for alcohol). There- fore, plausibility checks are carried out by comparing the data on private consumption from the national accounts and taking that data for the (detailed) weighting of the relevant index items. Other key sources of data are administrative institutions (e.g. the Austrian Association of Social Insurance Providers for health care data, and agencies supplying tourism and car registration statistics), microcensus surveys providing data on costs for shelter and surveys by market research institutes and polling organizations. Finally, the CPI also includes expenses for assisted living (e.g. retirement homes). Those expenses are not derived from consumer surveys but come from other sources (social security funds).

50 OESTERREICHISCHE NATIONALBANK The measurement of inflation in Austria: a historical overview

5 Ambiguities in inflation received particular attention. The re- measurement in light of the port was entitled “Toward a More Ac- Boskin Report on the U.S. CPI curate Measure of the Cost of Living” 5.1 Causes of measurement and was presented to the U.S. Senate inaccuracies in the calculation of Finance Committee by a scientific inflation committee made up of renowned U.S. Computing the rate of inflation in- university professors led by Michael J. volves complex survey and calculation Boskin for the purpose of describing processes that represent potential measurement errors in the U.S. CPI. sources of error with regard to execut- According to the report, the inflation ing field work and price surveys, deter- rates reported for private consumption mining and updating the basket of in the U.S.A. at that time exceeded goods and services, deriving expendi- the “true” rate of inflation by 1.1 per- ture weights, selecting retail establish- centage points per year, with a fluctua- ments, establishing conventions for add- tion margin of between 0.8% and ing new products and handling quality 1.6%. The Boskin Report identified four changes, calculating accurate prices (in potential sources of error that could view of fictitious prices such as rebates), lead to overestimation of the inflation dealing with seasonal goods and other rate. In the case of the U.S., 0.6 per- particularly significant expense groups centage points were attributable to such as rent, and allowing for substitu- insufficient depiction of changes in the tion effects based on consumer demand quality of new products, 0.4 percent- (in the case of goods and retail estab- age points to substitution effects re- lishments). sulting from changes in buying behav- The general public has become much ior in the selection of products, and more aware of inflation measurement 0.1 percentage points to the selection since the mid-1990s, when central of retail establishments. banks began making price stability and Although the report’s findings were low inflation rates a key objective. A disputed among experts and cannot be low rate of inflation has always been transposed 1:1 to other countries (since one of the main criteria for admission different measurement concepts are to EMU. Studies carried out prior to applied), at the time they were cause monetary integration show that the infor intense discussion and resulted in flation measurement methods used up the initiation of similar studies.20 to that time could result in a distortion – predominantly a significant overstat- 5.2 Methods of minimizing sources ing – of the upward movement in of error prices. A more general description of how In this context, the Boskin Report measurement errors can be minimized (Boskin and Dulberger, 1996, 1997) when calculating inflation is given in

20 Studies were also carried out for Canada (1997) and the U.K. (1995); see Hoffmann (1998, p. 7) for an overview. The margin of measurement error was less than 1 percentage point per year for both countries at the time (Canada: 0.5 to 0.7 percentage points; U.K. 0.35 to 0.8 percentage points). A study by the Deutsche Bundesbank found that the overstating bias was around 3/4 of a percentage point for the German CPI (Hoffmann, 1998). The majority of that error was caused by difficulties in measuring quality. There are no scientific studies on measurement errors in the CPI inflation rate for Austria; cautious estimates (from experts at Statistics Austria) referred to 0.1 to 0.3 percentage points.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 51 The measurement of inflation in Austria: a historical overview

the following subchapter (for some top- goods in the index is a frequent lack of ics with special reference to current reliable, recent statistics documenting Austrian practice). their demand, which even if available may not allow for separation of the de- 5.2.1 Consideration of quality changes mand portion attributable to the com- Nearly all goods are subject to qualita- mercial sector and that attributable to tive changes resulting from changes in the household sector. Two legislative model, equipment or workmanship; changes have contributed to an im- price comparisons over time are only provement in measurement accuracy. useful for homogeneous goods. Due to First, baskets of goods and services are the inevitability of changes in the sup- now adjusted at shorter intervals than ply of goods and, increasingly, in ser- in the past, and second, the EU has leg- vices as well, price series for old and islated that any new goods representing new models (or old and new rates or 0.1% of consumer demand must be in- fees) must frequently be chained for cluded in the HICP. Statistical offices price calculation purposes. are required to use recent statistics to One of the most difficult practical conduct ongoing reviews of the signifi- considerations in the preparation of in- cance of new consumer products for dices is how to assess whether a subse- the price index. quent model is a higher-quality product or merely another product with the 5.2.3 Accounting for changes in demand same features (a “type change”) as well habits… as how to differentiate between price The Laspeyres effect describes a phenom- and quality changes and the resulting enon in which consumer demand in- impact on inflation. creases when a product registers a be- This is especially important for in- low-average increase in price, or even a dex concepts that follow the Laspeyres drop in price, and tends to decrease approach and, strictly speaking, always when goods become more expensive. It observe price trends for the same prod- was long not possible to account for uct. In such cases, the key consider- such changes in demand habits in the ation is which portion of a change in index given the necessity of compro- price is relevant to inflation and which mising between accuracy and practical- is not. Various methods have been de- ity in the constant observation of price veloped (Statistics Austria, 2011; ILO trends. This has become easier now et al., 2013; Eurostat, 2016) for the that baskets of goods and services are purpose of facilitating such assessments being adapted more frequently. The and have to be applied in the harmo- more often a basket is updated, the less nized calculation of inflation rates prevalent the substitution effect is. across the EU (box 3). 5.2.4 ... and for switching retail 5.2.2 Timely inclusion of new goods and establishments services In Austria, local authorities in the 20 A problem arises when new goods have cities reporting price information are already achieved a certain market share responsible for the selection of report- but do not appear in the index, despite ing retail establishments. The obliga- having caused drastic price reductions tion of the establishments to cooperate once they entered into mass produc- in price sampling is set forth in Article tion. The difficulty in including such 7 of Regulation (EU) No 2494/1995

52 OESTERREICHISCHE NATIONALBANK The measurement of inflation in Austria: a historical overview

Box 3

Methods of adjusting for quality in the Austrian CPI and HICP In Austria, a committee of experts (CPI working group)1 offers advice and decides on a case-by- case basis on whether quality changes or changes in type have occurred. Application of the basic rule – “a quality change exists if an essential product feature has changed, and a type change exists if only accessory features (that are insignificant for the price) have changed” – is not always possible, however. Making an objective assessment of quality changes requires detailed discussion as well as various aids (surveys of sector experts, inquiries to manufacturers, decomposing the item into its price-determining features). The database set up by Eurostat, which specifies criteria and provides examples, serves as another aid in decision making. A subjective assessment of changes in quality, i.e. a consumer assessment and an assessment of the benefits gained – although likewise theoretically significant – is difficult to measure statistically and must rely on features such as improved safety, more efficient application, and the like. In addition to direct price comparisons (where no essential quality features have changed), both implicit and explicit methods are used to make quality adjustments to the Austrian price index (Statistics Austria, 2011). Implicit methods measure changes in price between two conse- cutive price observation points on the basis of price changes in similar models during the same period. Explicit methods assess quality differences using observable differences in product features; these are the methods most frequently applied to calculation of the Austrian CPI and HICP. The following is a simplified explanation of the types of explicit quality assessment methods used:

Option cost method (with 50% rule): Substitute item B includes a feature that could only be purchased as an option separately from item A in period t. The quality adjustment between item A and item B equals half of the price of the separate option.

Expert assessment: Manufacturers, consumers and experts make a subjective, graduated estimate of the quality change and degree of price change.

Hedonic calculations: A regression model in which the item is broken down into its consti- tuent features serves to determine which features are responsible for the price difference.

The option cost method is used in Austria e.g. for cars and personal computers. The hedonic method is used for books and SD memory cards. In addition, expert assessments from the CPI working group are often taken into account. Since the figures calculated are only an indication of quality changes rather than exact monetary amounts, they are only considered in the context of a larger system. Price differences are broken down into five levels ranging from no quality difference (type change S, or Q0), Q1 (one-quarter of the price difference is quality related), Q2 (one-half of the price difference is quality related), and Q3 (three-quarters of the price difference are quality related) to the price difference equaling the quality difference (Q4). Thus either the entire price difference, a portion of the difference or no price effect are factored into the inflation rate. In Austria, quality adjustments are made for approximately 3% of all goods and services each year. With regard to the implicit methods, overlap pricing is of some significance when substituting products in the context of a complete revision of the basket of goods and services. Overlap pricing is used when item A and substitute item B are both on the market for at least one period t (overlap period), with the quality adjustment of B in the period t+1 (when only B is available) being equal to the price difference between A and B in period t (Statistics Austria, 2011).

1 A Statistics Austria committee composed of experts from various institutions (ministries, Chamber of Labour, Austrian Economic Chambers, WIFO, OeNB).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 53 The measurement of inflation in Austria: a historical overview

and in Article 5 of the new Regulation 5.2.5 Eliminating sources of errors in (EU) 2016/792, which for the first field work time requires the provision of scanner The majority of price data comes from data where available. Regional selection survey institutions, and the field work committees endeavor to acquire the carried out to collect the price data is most market-relevant sector represen- prone to error. However, sampling tatives for the sampling. However, errors are not likely to be especially sometimes retail establishments are re- prevalent among those official price sistant to supplying a price-surveying data collectors using computer-assisted organization with information on the technologies (Austria is currently in the prices quoted to each customer. Thus it process of transitioning to tablets for does not appear to be generally assured price data collection). Where data are that market leaders or discounters whose collected by temporary staff that have value for money or price adjustment received training in collection method- strategies and price trends differ from ology, but have less experience, the those of other retail establishments will provincial CPI administration and be included in the CPI and HICP. Statistics Austria have to ensure that Moreover, since total revenues are not the downstream control mechanisms used as a price-weighting factor, the and plausibility checks are sufficient to price of an item is calculated from identify any inaccuracies. the unweighted average of all prices reported. It is therefore of particular 6 EMU: harmonized inflation relevance that a representative selection measurement becomes state of of retail establishments and types of the art establishment be made. A low rate of inflation has always been one of the main convergence criteria

Chart 1 CPI and HICP inﬂation rates in Austria % 4.0

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0.0 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 Feb. 16 Apr. 16 HICP CPI

Source: Statistics Austria.

54 OESTERREICHISCHE NATIONALBANK The measurement of inflation in Austria: a historical overview

Box 4

HICP inflation measurement governed by more than 20 EU regulations, but a consolidated legal basis is being implemented In 1994, an EU-level working group was set up to develop a harmonized consumer price index. The working group was composed of representatives from all statistical offices plus selected central banks and specialists in the field. In 1995, Council Regulation (EC) No 2494/95 was enacted. The regulation governs the objectives and definition of the HICP and lists comparability requirements in addition to setting the initial minimum standards for index criteria and specifying the procedure for publishing the HICP across the EU. Another 22 regulations (Euro- stat, 2016a) cover the entire methodology by refining the preparation and calculation methods specified in the framework regulation; certain aspects of HICP preparation (e.g. weighting) were specified multiple times over the years based on practical experience gathered during HICP preparation. Moreover, the new Regulation 2016/792 and the implementing regulation to be adopted in 2017 consolidate and update the HICP-relevant regulations in effect today. The regulations are supplemented by recommendations, e.g. on rents, telecommunications, and health care (Eurostat, 2016b). They describe procedures for preparing the HICP and go beyond the statutory requirements, so they are not legally binding. The original HICP calculations did not include data on owner-occupied housing, which is particularly difficult to compare due to the different methods by which the national CPIs deal with this item. Treatment of owner-occupied housing was not legally regulated until 2013, when Regulation (EU) No 93/2013 defined criteria for creating price indices for owner-occupied housing. The price index for owner-occupied housing tracks price trends in all direct and indirect components of real estate transactions and therefore supplements the HICP as a key indicator for estimating price trends at the consumer level.1 In addition to the main index, there are two separate subindices: “Purchase of new homes” and “Purchase of existing housing stock.” The primary factors relevant to preparation of the Austrian index are presented in sections 4.6 and 4.7, which also discuss selected, legally binding aspects of HICP index preparation as well as the differences between calculating the HICP index and the national CPI.

1 The European Commission will publish a report by end-2018 on the inclusion of this index in the HICP (Regulation (EU) No 2016/792, recital 10, p. 2). for Member States to participate in developed EU-wide guidelines for im- EMU. Many aspects of inflation mea- proving comparability as well as help- surement in the EU Member States ing to minimize possible sources of were nonhomogenous (methodology, error (Fluch, 1999; Hackenberg-Vögl, collection, determination of the basket 1999). This gradually led to the devel- of goods and services, delineation of opment of an HICP manual based on consumption groups e.g. with respect EU regulations that serves as a theoret- to the inclusion of owner-occupied ical and practical guide for measuring housing), meaning that the inflation inflation in the EU Member States rates calculated were only comparable (Eurostat, 2016c). to a limited extent. For this reason, a Ever since the start of monetary working group headed by Eurostat was union, the HICP has acted as the pri- established at the beginning of the mary indicator of inflation for the ECB 1990s and tasked with creating an and the Eurosystem in setting mone- HICP for the EU (box 3). The working tary policy. The HICP is a key part of group later began incorporating find- monetary policy strategy and provides ings from the Boskin Report and other information on the stability of the euro. studies into the HICP concept and The Eurosystem aims to maintain the

MONETARY POLICY & THE ECONOMY Q3– Q4/16 55 The measurement of inflation in Austria: a historical overview

annual HICP inflation rate at near or 1921 that official statistics reporting below 2% in the euro area. began in the wake of high inflation or The calculation and publication of hyperinflation in Austria. Various indi- inflation rates and the categorization of ces were calculated (predominantly for consumption groups and subgroups Vienna only), with backward calcula- ensures that the information on price tions extending back to 1914 in some trends is sufficiently detailed and cur- cases. The historical series from 1800 rent. HICP statistics also allows mone- to 1914 are based on backward calcula- tary policymakers to undertake current tions from the late 1970s, which were analyses of core inflation indicators prepared using 19th century price re- (total inflation excluding volatile goods cords for Austria’s main cities and ap- such as fuels, seasonal foods and tour- plied a household spending model simi- ism services) and to prepare monthly lar to that used in 1912 for the first flash estimates as well as projections of modern consumer survey. In the inter- euro area inflation. war period as well as in the years up to Now, after approximately 20 years, 1959, various types of cost-of-living in- the HICP has achieved recognition as a dices and retail price indices were com- reliable tool for measuring the inflation puted to serve as parameters for changes rate in the EU and the euro area. As in inflation. Two World Wars, severe mentioned above, the methods for cal- living conditions for the population, culating the HICP have been applied to state control (price fixing, rationing) some aspects of the calculation of na- and repeated currency substitutions tional CPIs, which continue to exist in posed great challenges for the calcula- nearly all EU Member States, including tion of price indices at that time. CPIs Austria. Some conceptual differences similar to today’s HICP – though less between the HICP and the CPI (see sec- sophisticated in their methodology and tion 4.6 and Fluch and Rumler, 2005) measurement accuracy – were not remain unchanged, but cause only min- available until the 1958 consumer price imal deviations of 0.1 percentage point index. The CPI 58 was prepared by the to 0.2 percentage points in the inflation Austrian Central Statistical Office rates depicted by the two indices for based on the spending patterns of all Austria (Auer and Pesendorfer, 2016). Austrian households of the 1955 consumer survey. 7 Summary and outlook In the 15 generations of CPIs calcu- A CPI measures changes in the prices of lated for Austria since 1800 (and HICPs goods and services that households con- since 1996), drastic changes have oc- sume. The calculation of price indices curred in household consumption pat- at international level began as early as terns, the composition of the basket of the beginning of the 18th century. In goods and services, the criteria for Austria, consumer price indices have price data collection and index con- been consistently available since 1800. struction methods, all of which are key Though simple, the indices followed to- elements of a CPI. Whereas the first day’s basic structure, and chaining the price indices were still characterized by different price indices allows for con- the predominance of basic necessities, clusions to be drawn on inflation rate consumption patterns are vastly differ- trends, adjustments for inflation and ent today. The proportion of basic ne- indexation for a period now spanning cessities has declined significantly and more than 200 years. It was not until luxury items have increased to account

56 OESTERREICHISCHE NATIONALBANK The measurement of inflation in Austria: a historical overview

for some 75% of the basket of goods tion rates by the central banks, innova- and services in 2016. The coverage of tive index producers, critical use of the basket was also greatly expanded. consumer price indices, the accompa- The first international scientists dealing nying scientific work on price indices with price indices used only very few and the use of comprehensive statistical (three to five) goods, and the 1922 instruments. Today’s calculations of COLI included 40 goods. By 1959, the inflation rates in the form of national basket already included 197 items, and CPIs and HICPs have without a doubt by 2000 the number of goods and ser- reached a high level of acceptance vices in the basket had quadrupled. among monetary and economic policy- Since then, the number of items in makers due to their high quality, quick the basket has been relatively stable at availability and the disaggregated data around 800. preparation. The same applies to the construc- Inflation measurement will none- tion of the respective generations of theless continue to face the following indices. The historical indices from the challenges: 19th and early 20th century include • The fast pace of product cycles and 13,000 individual prices for the calcu- heterogeneity of offers on the market, lated years from 1800 to 1914. By 1959, with their complex pricing packages some 5,000 prices were being recorded and fee systems, will demand more each month, and in 2016, the CPI and attention than in the past with re- HICP are calculated on the basis of ap- spect to making quality adjustments. proximately 40,000 individual prices. • The increase in online shopping by A similar momentum was seen in the consumers and rapid price adjust- retail establishment samples drawn in ments due to technical integration – the same period with an increase from daily or even hourly pricing (dynamic approximately 1,300 to more than system pricing) in online shops – will 4,000. Computer-assisted processing alter price data collection methods and control mechanisms have replaced such that price registration systems complicated procedures for maintain- will have to become more flexible ing price lists and manual processing, rather than relying on just one or two and tablets are now being used to facili- price data collection dates each tate field work. month. This will necessitate addi- As shown in various studies, mea- tional field work. surement distortions affecting the accu- • The massive increase in customer racy of inflation calculations can arise loyalty programs has resulted in ficti- from the following index factors: the tious prices as well as constant sale frequency of weight adjustments, the prices in nearly all consumer seg- tempo at which new goods achieve mar- ments – this having become the rule ket and demand relevance for house- rather than the exception – will holds and are included in the calcula- require strategies for identifying the tion of inflation, and the precision with most frequent prices actually paid which quality changes are included. by customers and including them in The quality of inflation measurement the sample. It remains to be seen if has improved steadily, however, thanks Statistics Austria’s transition to using to the findings from the Boskin Report, scanner data for some sectors, which extensive work in the EU to harmonize is currently in the preparatory stages, the HICP, the precise analysis of infla- will succeed in supplying price data

MONETARY POLICY & THE ECONOMY Q3– Q4/16 57 The measurement of inflation in Austria: a historical overview

that are more realistic than the dis- tion rates that better corresponded to played prices or list prices currently the perceived prices for household used for inflation measurement. items. The calculation of individual • The preparation of special indices has rates of inflation for households21 proven useful and should be contin- using special tools is another innova- ued. Back between 2002 and 2005, tion worth mentioning, since this when the published inflation rate di- allows consumers to estimate the verged from the perceived rate of in- changes in their personal buying flation after introduction of euro power. This can help alleviate the cash, price statisticians in Austria de- criticized discrepancy between mac- veloped special baskets of goods and roeconomic inflation and individual services and indices capable of mak- perceptions of inflation. ing more precise calculations of infla-

References Arbeiterkammer Wien (Vienna Chamber of Labour). 1955. Wirtschaftsstatistisches Jahrbuch. Antonowicz, W., E. Dutz, C. Köpf and B. Mussak. 2016. Memories of a central bank. Oesterreichische Nationalbank. Since 1816. Brandstätter. Auer, J. and K. Pesendorfer. 2016. Inflation im Jahr 2015 – Analysen und Trends. www.statistik.at/ web_de/statistiken/wirtschaft/preise/verbraucherpreisindex_vpi_hvpi/index.html (retrieved on June 30, 2016). Boskin, J. M. and E. R. Dulberger. 1996. Toward a More Accurate Measure of the Cost of Living. Final Report to the Senate Finance Committee from the Advisory Commission to Study the Consumer Price Index. In U.S. Senate Finance Committee. 104th Congress. 2nd Session. Senate Print. Washington, D. C. December. 104–172. Boskin J. M. and E. R. Dulberger. 1997. The CPI Commission: Findings and Recommenda- tions. In: The American Economic Review 87(2). 78–83. Bundesamt für Statistik (Federal Statistical Office). 1922. Mitteilungen des Bundesamtes für Statistik 1. Bundesamt für Statistik (Federal Statistical Office). 1925–1938. Statistische Nach- richten. Several issues. Cvrcek, T. 2013. Wages, Prices, and Living Standards in the Habsburg Empire, 1827–1910. In: The Journal of Economic History 73(01). Cambridge University Press. March. 1–37. Diewert, W. E. 1988. The Early History of Price Index Research. In: NBER Working Paper Series. Working Paper 2713. September. Emmerig, H. 2012. Preise und Löhne in Österreich. Rechnungen und Rechnungsbücher vom 12. bis 20. Jahrhundert. Mitteilungsblatt 44. Universität Wien. Institut für Numismatik und Geldgeschichte. Update January 2015. Eurostat. 2016a. Legislation. http://ec.europa.eu/eurostat/web/hicp/legislation (retrieved on August 3, 2016). Eurostat. 2016b. Recommendations. http://ec.europa.eu/eurostat/web/hicp/methodology/ recommendations (retrieved on August 3, 2016). Eurostat. 2016c. HICP methodology. http://ec.europa.eu/eurostat/statistics-explained/index. php/HICP_methodology (retrieved on August 3, 2016.).

21 Statistics Austria offers a tool for calculating personal inflation rates: www.statistik.at/persoenlicher_ inflationsrechner.

58 OESTERREICHISCHE NATIONALBANK The measurement of inflation in Austria: a historical overview

Fluch, M. 1999. Harmonized Indices of Consumer Prices – Progress and Unresolved Problems in Measuring Inflation. In: Focus on Austria 2/99. 54–73. Fluch, M. and F. Rumler. 2005. Unterschiede des VPI und HVPI in Österreich – Sind zwei parallele Verbraucherpreisindizes zweckmäßig. In: Statistiken Daten & Analysen Q4/05. 67–77. Fluch, M., F. Rumler and T. Wittenberger. 2010. Discounting Pricing in Austria: Insights into Retail Business Practices and HICP Coverage. In: Monetary Policy & the Economy Q3/10. 59–89. Hackenberg-Vögel, I. 1999. Der Harmonisierte Verbraucherpreisindex als neuer Indikator für die Geldpolitik des Europäischen Systems der Zentralbanken in der 3. Stufe der Wirtschafts- und Währungsunion. Diploma thesis. University of Economics and Business: Vienna. Hoffmann. J. 1998. Probleme der Inflationsmessung in Deutschland. In: Diskussionspapier 1/98 Volkswirtschaftliche Forschungsgruppe der Deutschen Bundesbank. ILO, IMF, OECD, Eurostat, UN and Worldbank. 2013. Consumer price index manual – Theory and Practice. Jobst, C. and H. Kernbauer. 2016. The Quest for Stable Money. Central Banking in Austria, 1816–2016. Campus. Frankfurt. New York. Klezl, F. 1925. Die Lebenskosten. In: Geldentwertung und Stabilisierung in ihren Einflüssen auf die soziale Entwicklung in Österreich 169. Band der Schriften des Vereins für Sozialpolitik. Mühlpeck, V., R. Sandgruber and H. Woitek. 1979. Index der Verbraucherpreise 1800–1914. Geschichte und Ergebnisse der zentralen amtlichen Statistik in Österreich. Eine Rückrechnung für Wien und den Gebietsstand des heutigen Österreichs 1829–1979. In: Österreichisches Statistisches Zentralamt. Beiträge zur österreichischen Statistik 550. OECD. 1997. Synthesis Paper on Shortcomings of the Consumer Price Index Measure of Inflation for Economic Policy Purposes. Paper prepared for the Working Party No 1 on Macroeconomic and Structural Policy Analysis. Österreichisches Statistisches Zentralamt (Austrian Central Statistical Office). 1977. Index der Verbraucherpreise – Revision 1976. Österreichisches Statistisches Zentralamt (Austrian Central Statistical Office). 1979a. Beiträge zur österreichischen Statistik 550. Geschichte und Ergebnisse der zentralen amtlichen Statistik in Österreich 1829–1979. Österreichisches Statistisches Zentralamt (Austrian Central Statistical Office). 1979b. Beiträge zur österreichischen Statistik 550A. Annex of tables to 550. Österreichisches Statistisches Zentralamt (Austrian Central Statistical Office). 1987. Beiträge zur österreichischen Statistik 853. Verbraucherpreisindex – Revision 1986. Österreichisches Statistisches Zentralamt (Austrian Central Statistical Office). 1990. Beiträge zur österreichischen Statistik 956. Entwicklung der Verbraucherpreise seit 1900. Österreichisches Statistisches Zentralamt (Austrian Central Statistical Office). 1997. Verbraucherpreisindex 96 und HVPI -Gewichtungsschemata. In: Statistische Nachrichten 5. 356 –367. Österreichisches Statistisches Zentralamt (Austrian Central Statistical Office). 1998. Indizes zur Wertsicherung. Anleitungen, Langzeitreihen. Beispiele. Österreichisches Statistisches Zentralamt (Austrian Central Statistical Office) and WIFO – (Austrian Institute of Economic Research). 1959. Neue Indizes der Verbbraucherpreise. Addendum to March issue of Statistische Nachrichten XIV(3) and (12). Special issue. Pribram, A. F. 1938. Materialien zur Geschichte der Preise und Löhne in Österreich. Vienna. Statistics Austria. 2001. Der neue Verbraucherpreisindex 2000 – nationaler und harmonisierter Verbraucherpreisindex.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 59 The measurement of inflation in Austria: a historical overview

Statistics Austria. 2011. Verbraucherpreisindex und Harmonisierter Verbraucherpreisindex ab 2011. www.statistik.at/web_de/statistiken/wirtschaft/preise/verbraucherpreisindex_vpi_hvpi/ index.html (retrieved on July 12, 2016). Statistics Austria. 2013. Konsumerhebung 2009/10. Standard-Dokumentation – Metainformationen. Statistische Zentralkommission (Central Statistical Commission). 1921. Mitteilungen der Statistischen Zentralkommission 1. Suppanz, C. 1976. Die österreichische Inflation 1918–1922. Forschungsbericht/Research Memo- randum 111. November. WIFO (Austrian Institute of Economic Research). 1947. Lebenshaltungskosten. In: Monats- berichte 1/3 1947. 12–14. WIFO (Austrian Institute of Economic Research). 1949. Zur Problematik des Lebens haltungskostenindex. In: Monatsberichte 1/1949.

60 OESTERREICHISCHE NATIONALBANK Two centuries of currency policy in Austria

This paper is devoted to currency policies in Austria over the last 200 years, attempting to Heinz Handler1 sketch historical developments and uncover regularities and interconnections with macroeconomic variables. While during the 19th century the exchange rate resembled a kind of technical relation, since World War I (WW I) it has evolved as a policy instrument with the main objectives of controlling inflation and fostering productivity. During most of the 200-year period, Austrian currencies were subject to fixed exchange rates, in the form of silver and gold standards in the 19th century, as a gold-exchange standard and hard currency policy in much of the 20th century, and with the euro as the single currency in the early 21st century. Given Austria’s euro area membership, national exchange rate policy has been relinquished in favor of a common currency which itself is floating vis-à-vis third currencies. Austria’s predilection for keeping exchange rates stable is due not least to the country’s transformation from one of Europe’s few great powers (up to WW I) to a small open economy closely tied to the large German economy. JEL classification: E58, F31, N13, N14, N23, N24 Keywords: currency history, exchange rate policy, central bank, Austria

When the privilegirte oesterreichische versus flexible exchange rates. During National-Bank (now Oesterreichische most of the period considered here, Nationalbank – OeNB)2 was chartered Austrian currencies were subject to in 1816, the currency systems of major fixed exchange rates, in the form of sil- nations were not standardized by any ver and gold standards in the 19th cen- formal agreement, although in practice tury, as a gold-exchange standard in a sort of specie standard prevailed. much of the 20th century and with the Functioning bullion markets and the euro as the single currency in the early absence of capital restrictions produced 21st century. In between, brief phases of roughly constant bilateral exchange exchange rate flexibility prevailed (e.g. rates of participating gold or silver cur- following the end of the Bretton Woods rencies. With the introduction of paper system), but before long they would money and the practice of coinage give way to another peg system, includ- debasement, financial breakdowns often ing Austria’s renowned hard currency mirrored the abuse of the currency sys- policy and, later, euro area membership. tem for government financing purposes. Section 1 briefly reviews the possible Measurable changes in currency prices role of the exchange rate as an eco- were overwhelmingly the result of wars nomic policy instrument. The subse- and revolutions and their impact on quent two sections, which rely heavily government finance and inflation. on Jobst and Kernbauer (2016) and Repeated crises, which were already Butschek (1985), are devoted to the occurring during the gold standard development of currency relations in period of the late 19th and early 20th Austria and their linkages to the rest of century and which were aggravated by the economy: Section 2 deals with the the experience of the unstable interwar era before World War I (WW I) and period, sparked the discussion on fixed section 3 with exchange rate policies

1 Vienna University of Technology, Institute of Statistics and Mathematical Methods in Economics and Emeritus Consultant, Austrian Institute of Economic Research (WIFO), [email protected]. The author would like to thank Ernest Gnan, Maria Teresa Valderrama (both OeNB) and the referee for helpful comments and valuable Refereed by: suggestions. Of course, responsibility for remaining flaws lies with the author. Niels Thygesen, 2 In this contribution, “OeNB” will be used as an abbreviation for the various names the OeNB had in the course of University of its 200-year history, including the Austro-Hungarian Bank. Copenhagen

MONETARY POLICY & THE ECONOMY Q3– Q4/16 61 Two centuries of currency policy in Austria

and their effects after WW I and up to An early form of currency policy the present day. Section 4 examines occurred when paper money and coin- the interaction of exchange rates and age debasement eroded the value of a other economic policy variables such as currency, with the effect that the in- inflation, fiscal balances and external trinsic value of a coin or banknote accounts. Section 5 summarizes. Why sagged below its face value. When a choose WW I as the dividing line be- currency’s link to specie gets lost, the tween sections 2 and 3? This is because currency’s credibility must be fur- in the earlier period, exchange rate nished by the economic fundamentals policy was confined to managing cur- and credible economic policies of the rency relations around essentially fixed issuing country, or at least by binding parities, whereas thereafter it has been legal restrictions on its asset holdings. utilized as an active element in steering Currencies then become national enti- the macroeconomy. Also, before WW I ties, and the exchange rate a measure of the dominating policy goal was to the relative economic strength of two balance the external accounts, while nations. With the varying importance thereafter internal targets such as eco- of specie coins, paper money and nomic growth and full employment foreign exchange, various monetary came to the forefront. regimes have evolved: • bimetallism in the early 19th century; 1 The exchange rate as a policy • the classical gold standard in the late tool 19th and early 20th centuries; Exchange rate policy in the modern • a variety of fixed and floating ex- sense gained importance only after the change rate regimes in the turbulent decline of the classical gold standard interwar period; during WW I. In previous times, vari- • the Bretton Woods system of gold- ous precious metals had been used as based fixed but adjustable exchange currencies, often regardless of national rates; and borders. The prices of silver and gold • generalized floating thereafter, which were formed on the world market and, also comprises the era of regional together with the gold and silver con- monetary integration in Europe. tent of a regional coin, they defined the These periods may also be differenti- price of that coin in terms of another ated by their regime of capital restric- region’s coin. Such relative prices were tions: Free capital movements had a independent of considerations concern- limited, though growing, influence ing the economic position of the re- during bimetallism and the gold stan- spective region. In fact, the metal standard and have been a typical character- dard acted as an external anchor for the istic of the post-Bretton Woods man- currency. Currency prices represented aged floating system, while capital con- technical relations; they were not (yet) trols prevailed during the interwar utilized as policy instruments. Public period and in the early years of the authorities were involved, though, when Bretton Woods system. a currency’s parity was determined, Under the heading “exchange rate i.e. the number of face-value coins to policy,” two theoretical strands may be minted from, say, a pound of silver, be investigated: (1) determinants of or later on the number of silver ounces exchange rate changes (the flexible ex- to be exchanged for a unit of paper change rate as endogenous variable); (2) currency. expected consequences of steering the

62 OESTERREICHISCHE NATIONALBANK Two centuries of currency policy in Austria

exchange rate for the rest of the econ- dynamics and the eventual impact omy (the exchange rate as exogenous on domestic inflation depending on variable). the pass-through effect. A dominant When investigating exchange rate objective of pegging the exchange determination in the long run (short- rate to a stable anchor – be it a pre- run variations are not the focus of the cious metal or a foreign currency – is current paper), major textbook expla- thus to control price developments. nations, which are not mutually exclu- During the silver and gold standard sive, are: periods, the aim was to preserve • The purchasing power parity (PPP) credibility and the value of the cur- theory, which asserts that, in the ab- rency. In the case of Austria’s hard sence of market distortions, the currency policy, the goal was to sup- (floating) nominal exchange rate be- press cost increases and thereby also tween two currencies reflects the induce advances in productivity. relative price levels (of traded goods) • Via the terms-of-trade effect, a cur- in the two countries in question; for rency depreciation is perceived to changes over time it means that a contain or reduce a deficit in the cur- country with comparatively high in- rent account, with price elasticities flation would experience a depreciat- of import and export demand deter- ing currency. mining the dynamics (J-curve effect) • Relative prices are often just mirror- and final result (Marshall-Lerner ing the underlying monetary condi- condition). Textbook theory also tions, with the implication that under considers the effects of this result on a floating regime excessive money domestic spending behavior and supply will fuel inflation and gener- capacity constraints (absorption ap- ate a currency depreciation. If long- proach) as well as monetary condi- term capital transactions are also con- tions. sidered (portfolio balance approach In practice, exchange rates have at dif- to exchange rate determination), a ferent times been used as a policy tool relative decrease in domestic long- for different purposes. Dominant goals term interest rates will entail capital of steering nominal exchange rates have outflows and currency depreciation. been to dampen exchange rate variabil- • With regard to the real sector of the ity and to diminish price increases, as economy, structural differences be- was the case during Austria’s hard tween two countries may comple- currency policy period. Real exchange ment the price effects and determine rates can only be targeted indirectly the development of real exchange through nominal exchange rates and rates. Relative increases in home pro- price developments. Real depreciation ductivity or shifts in consumer pref- would contribute to improving com- erences from foreign to domestic petitiveness and thereby reduce a defi- goods would thus tend to appreciate cit in the current account. In the short the real exchange rate. run, changes in exchange rates have of- Turning the relation on its head, what ten been used as shock absorbers to are major long-run channels of ex- mitigate the impact of external shocks change rate changes to the rest of the on the real sector of the economy. This economy? function requires that changes in nomi- • A change in the exchange rate imme- nal exchange rates eventually show up diately alters the terms of trade, the as changes in the real exchange rates.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 63 Two centuries of currency policy in Austria

With respect to empirical realiza- had climbed to some 500% by 1811. tions, the theoretical hypotheses sug- Eventually, national bankruptcy was gest that in the long run one-way rela- declared in 1811 and a currency reform tionships are rare, and one will rather initiated. Bancozettel were devalued by observe a co-movement of exchange 80% and had to be converted into rates and other economic variables. Wiener Währung, a new government However, the relationships mentioned paper money. Although the amount of here are setting the stage for the fol- Wiener Währung in circulation was lowing historical overview and for the restricted, the reform was jeopardized empirical investigation in section 4. by another new paper currency, the “Antizipationsscheine,” which were issued 2 Currency relations prior to from 1813 to anticipate the expected WW I revenues from a new land tax. In effect, The transition to the 19th century is too many Antizipationsscheine were characterized by a gradual formation of put in circulation without prior com- national currencies, linked together by munication to the general public, international prices of precious metals thereby eroding confidence in the ex- and liberal financial markets. Still mir- pected results of the currency reform. roring to some extent mercantilist ideas, To restore convertibility of Wiener economic policy at this time was largely Währung into Conventionsmünze, in 1816 oriented toward keeping the balance of the newly-founded OeNB was endowed payments in order. The prevailing spe- with a monopoly for printing banknotes cie standards secured constant bilateral and with a certain degree of indepen- exchange rates, as price fluctuations dence from the state administration, in the bullion markets did not alter although state financing was not ex- the relative currency values. Exchange plicitly prohibited. As stated by Jobst rates did thus not play a significant role and Kernbauer (2016: 52), “retiring as a policy tool. It was only the erratic, government paper money and replacing though gradually increasing, emission it with banknotes convertible into silver of paper money as an easy means of war was the primary objective of the authors and budget financing that provided a of the decree of 1816.” link to politics via inflation. Exchange The currency reform took many rate policy was confined to keeping years and was barely completed when fluctuations in currency prices low. the revolution of 1848 swept over to Austria. The banking system collapsed, 2.1 Early 19th century: War financ- and the OeNB covertly stepped in to ing, paper money and national finance the government by printing bankruptcy new money. When this was made pub- During the Napoleonic wars and the lic, a run on silver coins again ended subsequent war reparations, the origi- convertibility, initiating a law which nal redeemability of Austrian paper declared paper money as legal tender. money (Bancozettel) in silver coins The printing press was repeatedly acti- (florin Conventionsmünzen) was re- vated, as the Habsburg empire aspired voked. Between 1804 and 1808, the to be counted among the big European circulation of paper money multiplied players. However, neither the Austrian and massive inflation occurred. The economy nor the military machinery small silver premium against Banco would support such ambitions. As a re- zettel widened rapidly after 1806 and sult, the silver premium of the florin

64 OESTERREICHISCHE NATIONALBANK Two centuries of currency policy in Austria

fluctuated widely, peaking at some silver coins. As a consequence, the cen- 50% in the course of the year 1860. tral bank was swamped with silver ingots for coining, which induced the 2.2 From silver to gold decision to terminate the specie cover- The florin Conventionsmünze (fl. CM) re- age of currency altogether. In the years mained in circulation until 1857, when to follow, until the first half of the the Habsburg monarchy joined forces 1890s, the florin was maintained as a with the “Deutscher Zollverein” (Ger- paper currency, which is considered an man Customs Union) to reduce cur- early appearance of a flexible exchange rency diversity by adopting the “Vere- rate (Flandreau and Komlos, 2001; instaler,” a standard silver coin to be Chaloupek, 2003). used as a common currency based on When the Austro-Hungarian dual the rules of a full silver standard. The monarchy was established in 1867, a changeover to the new “österreichische switchover from silver to the gold stan- Währung” (ö.W.) resulted in a minor dard was discussed. Crown gold coins devaluation (fl.1 CM = fl.1.05 ö.W.). had already been minted since 1857 The florin ö.W. remained legal tender and used mostly in cross-country pay- until 1901. ments, and in 1862 the OeNB had been To underpin the currency reform, a authorized to invest in gold as a basis transition from fiscal dominance (fi- for money in circulation. However, it nancing the state) to monetary domi- was not until 1892 that the dual monar- nance (monitoring the interest rate) chy changed from a silver parity to a was initiated by the “Plenersche Bank gold parity: 1 florin (100 kreuzer) of akte” or Bank Act (Jobst and Kern- the old silver currency was set equal to bauer, 2016). The Bank Act, effective 2 crowns (à 100 heller) of the new gold- from the beginning of 1863, strength- based currency. Although gold convert- ened the OeNB’s independence, rede- ibility was never fully implemented, fined the coverage of banknotes by the Austro-Hungarian Bank intervened either discount operations (for up to fl. in the foreign exchange market to stabi- 200 million) or silver (for the rest of lize the crown’s exchange rate against the banknotes in circulation), and set a gold standard currencies. As of 1896, tight timetable for the reduction of gov- this moved Austria-Hungary de facto ernment debt. to a gold-exchange standard (Yeager, Following the war defeat of 1866 1998). Flandreau and Komlos (2001) against Prussia, Austria again slipped consider the period from 1896 to 1914 into huge budget deficits, an increasing a successful early experiment in creat- tax ratio and an expansion of the money ing an exchange rate target zone with a supply. Speculative financial market credible parity and narrow, though not activities around the world fair in prespecified, margins. From 1910 on- Vienna eventually led to the collapse of ward, the OeNB was obliged, as an an- the stock market in 1873. The rapid ti-inflation device, to maintain a stable dissemination of the classical gold stan- exchange rate of the crown vis-à-vis dard caused the market price of silver other major currencies. With hind- to tumble so much that by 1879 a sight, this period resembles Austria’s discount situation emerged, i.e. paper hard currency policy of the late 20th money became more expensive than century.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 65 Two centuries of currency policy in Austria

3 Exchange rate policies after hyperinflation. The value of the paper WW I crown collapsed. A turnaround was At the end of WW I, the monetary brought about in October 1922 when, regime of the Austro-Hungarian mon- under the auspices of the League of archy collapsed, and economic policy Nations, the “Geneva Protocol” was had to be oriented toward new objec- signed, committing Austria to remain tives. External equilibrium consider- an independent state, to prepare for the ations retreated into the background, issuance of international bonds (the while stabilizing the internal economy “Völkerbundanleihe”) on the basis of by reducing public debt and inflation the old gold crown of 1914, and to and restoring production capacities commence a major institutional reform became top priorities. With the end of including a stop on government financ- the classical gold standard, the exchange ing via the central bank. The reform rate was also discovered as an economic program immediately stabilized the in- policy instrument, and countries began flation rate as well as the exchange rate. to experiment with a return to some It included the emergence in 1923 of form of gold standard on the one hand, the OeNB out of the former Aus- and with flexible exchange rates on the tro-Hungarian Bank, and a changeover, other. The value of a currency would as of March 1925, from the crown to no longer necessarily be linked to the the Austrian schilling (10,000 crowns amount of specie in the vaults of the = 1 Austrian schilling), which remained central banks, but more and more to pegged to gold (at 0.21172086 grams of the economic fundamentals of an econ- fine gold to 1 schilling). However, omy, such as internal and external sta- Austria did not formally join other bility and total factor productivity. countries (e.g. the U.S.A., the United Kingdom, France and Italy) in reviving 3.1 Interwar consolidation period: the prewar gold standard, but instead from soft to hard currency pursued an independent peg to gold, During WW I, financing for the ex- backed by foreign exchange controls. A ploding government expenditures of major task for the OeNB was to secure the Austro-Hungarian monarchy was a stable value for the currency com- largely acquired through war bonds. pared with other enduring currencies The resulting massive increase in the with the eventual aim of returning to money supply pushed up inflation and some variety of the gold standard (Jobst caused the exchange rate of the crown and Kernbauer, 2016). to depreciate. At the end of the war, The international turbulence on the Austro-Hungarian Bank attempted stock exchanges did not become mani- to maintain a common currency with fest in Austria until May 1931, when the other successor states of the former Creditanstalt went bankrupt. The dis- monarchy, but failed when Yugoslavia closure by Creditanstalt of heavy losses and Czechoslovakia started to stamp in 1930 triggered a bank run which their own banknotes. To avoid an un- could only be managed by massive controlled influx of unstamped paper liquidity assistance from the OeNB and crowns, Austria, in now starkly re- a government guarantee for all Credit duced territorial confines, had to fol- anstalt deposits. Inflation expectations low suit. The war-ridden economy con- increased, capital flight began, and the tinued to face enormous budget deficits OeNB lost almost all of its foreign and inflation which in 1921 turned into exchange. While the OeNB maintained

66 OESTERREICHISCHE NATIONALBANK Two centuries of currency policy in Austria

the gold parity of the schilling, the re- was terminated in 1936 (Kernbauer, introduction of foreign exchange con- 1991, Klausinger, 2006). trols3 in October 1931 sparked a gold In March 1938 Austria was occu- premium which had surged to some pied and integrated into the Third 28% at the end of 1931 and 33% by Reich. In a first step the German June 1932. The OeNB raised the dis- reichsmark was introduced as the new count rate step by step to 10% in early currency, and the schilling was ex- 1932. A full return to gold parity would changed at the rate of 1.5 schillings per have required massive price and cost reichsmark (revalued from previously reductions, which seemed out of reach 2.17 schillings per reichsmark). The (Kernbauer, 1991). Eventually, as ma- functions of the OeNB as a central bank jor countries abandoned the gold stan- were suspended and its gold reserves dard (starting with the United Kingdom transferred to the German Reichsbank. in September 1931), the schilling was The schilling/reichsmark relation has devalued against gold by 22% in July later been gauged appropriate with re- 1932 (März, 1990). spect to relative wages (Butschek, 1985) Irrespective of this rather short and relative prices (Kernbauer, 1991). phase of a weak exchange rate, the longer-term performance of the schilling 3.2 From a war-ridden economy to was quite remarkable and yielded it the the Bretton Woods system nickname “Alpendollar” (Alpine dol- During World War II (WW II), when lar). This was underpinned by continu- the Bretton Woods system was negoti- ing foreign exchange controls (in effect ated, there was an overwhelming feel- until 1935) and by rather moderate ing among participants that economic budget deficits. The shrinking money stability could only be restored via supply (M1 contracted from 4.0 billion some sort of gold-exchange standard. schillings in 1928 to 2.2 billion schil- In effect, participants designed a sys- lings in 1934) was largely due to the tem with fixed but adjustable exchange reluctance of commercial banks to ex- rates and capital controls, though with tend credit to the private sector. The a commitment to gradually render cur- OeNB perceived its monetary policy as rencies convertible for current account not restrictive, as the monetary base in- transactions, while “corner solutions” creased slightly and the discount rate (rigid fixing or freely floating rates) could be lowered again. When in 1936 were rejected. a number of countries (among them Austria needed almost a decade to members of the ailing “gold bloc,” overcome postwar inflation and fiscal France, Italy and Switzerland) devalued imbalances, and to restore trade links their currencies, Austria, in an attempt and convertibility of the schilling as to counter inflation expectations, ab- preconditions for joining the gold-based stained from a comparable measure and Bretton Woods system. Immediately focused instead on accompanying de- after WW II, the exchange rate of the flationary measures. Thanks to the per- schilling was arbitrarily set by the ceived stability of the schilling, financial occupation forces to an overvalued supervision by the League of Nations 10 schillings per U.S. dollar, and for-

3 The spreading banking crisis had led to the introduction of foreign exchange controls in Germany as early as in July 1931.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 67 Two centuries of currency policy in Austria

eign exchange transactions for exports subsidies and an increase in prices. The and imports had to be authorized. In reform was not completed until May parallel, “grey” market notations re- 1953, when the schilling was effectively sulted de facto in an undervaluation of devalued by establishing a unitary ex- the schilling which, however, was grad- change rate at the level of the premium ually eaten up by swelling inflation. rate. In a parallel development, Austria Exporters switched to compensation gradually relinquished the restrictive contracts and premium contracts, postwar foreign trade regime by gradu- thereby engendering a system of multi- ally following the liberalization efforts ple exchange rates. of the General Agreement on Tariffs Among the immediate efforts to and Trade (GATT) and the Organiza- stabilize the war-ridden economy were tion for European Economic Co-opera- a series of currency-related laws restor- tion (OEEC). In 1953 trade in goods ing the functions of the OeNB as of and services was liberalized, and this 1922, reducing money in circulation by move was trailed by a stepwise liberal- 60% and limiting the usage of the other ization of cross-country financial flows. 40% to purchases of vital goods only – The schilling only became convertible thereby also preparing for the change- for foreigners in 1959, with full con- over from the reichsmark to the schil- vertibility being achieved in 1962. ling. In spite of these measures, infla- When in 1955 the Federal Act tion surged from 26% in 1946 to 96% on the Oesterreichische Nationalbank in 1947. To achieve stabilization, a first (Nationalbankgesetz) was amended, wage and price agreement was negoti- the main objective of the OeNB’s schil- ated between the social partners in July ling policy was changed from introduc- 1947 (four more were to follow by ing a gold standard to preserving inter- 1951). In November 1947, as a precon- nal purchasing power and maintaining dition for access to the European external value in terms of stable foreign Recovery Program, a further reduction currencies. At the same time, the of money in circulation was accom- OeNB’s independence was strength- plished (exchanging 3 “old” schillings ened and its operational spectrum for 1 “new” schilling), which caused the enlarged to include open market opera- market premium over the “official” tions and minimum reserve require- schilling exchange rate to dwindle ments. rapidly. Inflation in the late 1950s and early In August 1948, Austria became a 1960s was moderate and not as high as member of the Bretton Woods institu- in other partner countries. As the real tions by pledging to ease foreign ex- economy flourished and the current change restrictions and abandon the account developed favorably, the parity dual exchange rate system. In Novem- of 26 schillings per U.S. dollar was ber 1949, the foreign exchange system never endangered until the gold con- was simplified with the effect that, in vertibility of the U.S. dollar ended in addition to the devaluation of the “offi- August 1971. cial” exchange rate (now at the “basic rate” of 14.40 schillings per U.S. dol- 3.3 Stepwise approach to a hard lar), a “premium rate” of 26 schillings currency policy per U.S. dollar (close to the market The collapse of the Bretton Woods sys- rate) was established. For many trans- tem turned out to be an extended and actions this was tantamount to a loss of painful experience, lasting from the

68 OESTERREICHISCHE NATIONALBANK Two centuries of currency policy in Austria

end of U.S. dollar convertibility (August The first oil price crisis and the in- 1971) to the approval of generalized ternational recession of 1975 showed floating (March 1973). Although flexi- up in the large current account deficits ble exchange rates relieved many coun- of 1976/77. As a remedy, fiscal policy tries from balance of payments strains, turned restrictive and the OeNB this period is also characterized by high changed its focus to stabilizing the de- exchange rate variability. To make up preciation of the real effective exchange for the loss of gold as an anchor, central rate (REER), an endeavor in which it bank independence was emphasized was aided in any case by the deteriorat- and inflation targeting adopted as a new ing exchange rate of the U.S. dollar. In state of the art for a stability-oriented nominal terms, the OeNB attempted monetary policy strategy. In the Euro- to secure the credibility of the schilling pean Union (EU) this revived the dor- by reducing the variability of the schil- mant project to create, at least for a ling/Deutsche mark relationship which, large number of Member States, the from mid-1981, was practically nil. euro area as a regional fixed exchange Afterward, when divergences in the rate bloc. The international financial economic development of Austria and crisis of 2008, and in its wake the euro Germany surfaced, exchange rate sta- area sovereign debt crisis, have once bility was bought via an interest rate again called into question the meaning- premium in Austria over Germany. fulness of fixed exchange rates for In the EU, the 1990s were largely countries with strongly diverging eco- devoted to achieving the preconditions nomic fundamentals. for taking part in the stepwise process To contain the drawbacks of float- of Economic and Monetary Union ing exchange rates in general, and to al- (EMU), among other things by com- leviate price increases in particular, pleting the liberalization of interna- Austria was among the first countries tional capital transactions (Nauschnigg, to adopt a “hard currency policy” 2003). The credibility of a stable ex- (Handler, 1989, 2007). Already in May change rate was thereby upheld and 1971, the monetary authorities had was tested only once, when in August swung from balancing the current ac- 1993 a speculative attack on the schil- count to dampening inflationary pres- ling was countered by the OeNB with- sures through revaluing the schilling by out major strain. Overall, the hard cur- some 5%. Anti-inflationary arguments rency policy helped Austria to achieve were also advanced when (1) the OeNB and maintain price stability without ob- invented a composite currency Indica- vious negative side effects on economic tor to define the benchmark for the growth and employment. schilling exchange rate in August 1971; (2) the schilling was leaning toward the 3.4 Smooth transition to the single European “snake” currencies in an European currency attempt to reduce exchange rate vola- At the beginning of 1995 Austria joined tility within the “tunnel” of the Smith- the EU, aspiring to participate in all sonian Agreement from April 1972; steps toward completing EMU. Adopt- and (3) the Deutsche mark replaced the ing the euro from its start in 1999 snake currencies as the guideline for meant that Austria would relinquish the schilling exchange rate from July the schilling in favor of becoming a 1976 (Schmitz, 2016). member of a large currency union comprising its most important trading part-

MONETARY POLICY & THE ECONOMY Q3– Q4/16 69 Two centuries of currency policy in Austria

ners. The single currency is equivalent 4 Interactions between exchange to an irrevocably fixed exchange rate rates and other policy variables and implies that balance of payments Subject to the availability of reliable strains now have to be borne via adjust- historical data, this section provides an ments in the internal economy. overview of 200 years of exchange rate The euro has become one of the development in Austria in relation to dominating currencies in international the development of prices, fiscal bal- financial markets with flexible ex- ances and public debt as well as mone- change rates vis-à-vis other currencies. tary and real sector variables. The euro’s exchange rate developments have been a function of the interna- 4.1 Long-term effective appreciation tional relations between major eco- in nominal and real terms nomic areas, most importantly between To capture the whole exchange rate Europe and the U.S.A. Since 1999, spectrum of a country’s trade relations, large exchange rate swings have oc- effective (trade-weighted) nominal and curred, with the euro appreciating by real exchange rates (NEER and REER) some 85% from June 2001 to July would have to be considered. An appre- 2008, and thereafter depreciating by ciation of the REER (i.e. a loss in some 32% up to November 2015. relative competitiveness) results from Exchange rate policy in the euro either an appreciation of the NEER or a area is a shared competency of the relative increase in domestic prices. As Eurosystem and the EU Council. The chart 1 reveals, since the end of the former handles the daily interventions Bretton Woods system and up to the on foreign exchange markets while the mid-1990s the schilling significantly latter is responsible for the general appreciated in nominal terms (more orientation of the exchange rate system than 100%) as well as in real terms after consulting the European Central (some 30%), reflecting predominantly Bank (ECB). The OeNB is a member of Austria’s hard currency policy with its the Eurosystem and has a seat on the explicit aim of wage and price modera- Governing Council of the ECB. tion and productivity gains. In the mid-

Chart 1 Nominal and real eﬀective exchange rates for Austria from 1960 to 2015 Index 1960=100 220

200

180

160

140

120

100

80 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 NEER REER (labor cost-adjusted)

Source: AMECO.

70 OESTERREICHISCHE NATIONALBANK Two centuries of currency policy in Austria

1990s this general trend came to an flation per se. Over the long period end when, in the run-up to the com- considered here, nonwar periods have mon currency, budget deficits were been characterized by broadly parallel reduced and the relative labor cost po- developments of moderate inflation and sition improved. Overall, the much reasonably stable exchange rates, while faster nominal appreciation indicates marked interruptions occurred during that deteriorating price competitive- wars (Napoleonic wars, 1866 Prussian ness stemming from the increase in the War, WW I, WW II) and in the con- NEER was partly compensated by a de- text of currency reforms. While until crease in relative costs and prices. Since the end of WW II inflationary periods the turn of the century, NEER and were often succeeded by deflation, in REER movements have been closely the era since no protracted deflation correlated, which means that internal has occurred in Austria (table 1). and international prices have developed For the old silver-based florin Con- virtually in tandem. ventionsmünze, fluctuations in the market price of silver resulted in deviations 4.2 Inflation and exchange rates of the market value of a silver coin from According to the monetary approach to its face value. When paper currency exchange rate determination, excessive began to play a role in money supply, a relative money supply will cause a silver premium expressed the relative currency to depreciate. It also raises credibility of silver coins against inflation expectations and over time in- banknotes and state notes. This was the

Table 1 History of inflation and exchange rates in Austria

Period Inflation history Inflation rate Exchange rate history

Average Standard annual deviation change in %

1812–24 Deflation Currency reforms, stable silver –4.8 6.5 premium after 1818 1825–44 Price stability 0.3 3.6 Stable silver premium 1845–63 Moderate inflation 3.0 5.6 Fluctuating silver premium 1864–73 Deflation compensated by Widely fluctuating silver premium subsequent inflation 0.7 4.1 1874–1904 Moderate deflation until mid-1890s –0.6 2.1 Silver discount, gold parity from 1892 1905–14 Moderate inflation 2.1 2.1 Gold parity maintained 1915–24 WW I and hyperinflation 381.5 878.2 Massive crown depreciation 1925–45 Virtual price stability Currency reform, gold parity of schilling upheld, from 1938 reichsmark 1.0 2.9 as new currency 1946–52 Post-WW II inflation Currency reform, multiple exchange 31.3 29.1 rates 1953–70 Moderate inflation Fixed exchange rate vis-à-vis the U.S. 2.8 1.5 dollar 1971–81 Austro-Keynesianism, inflation, oil Post-Bretton Woods uncertainty, price crises Austria turning to “hard” Deutsche 6.3 1.9 mark as anchor 1982–98 Moderate inflation 2.9 1.3 Hard currency policy 1999 to present Inflation at ECB target 1.9 1.1 Single European currency

Source: Author’s specifications.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 71 Two centuries of currency policy in Austria

case in Austria during the Napoleonic following currency reform which wars, when the original credibility of brought silver convertibility to the new paper money was lost due to an exces- Austrian florin (for the years 1856– sive use of the printing press. Thereaf- 1858). The premium turned to a dis- ter, the silver premium was closely count in 1879 and remained close to linked to external and internal political zero until 1886. Because of a dwindling turbulences and related developments silver price in terms of gold, the silver in government finances. In 1818, the discount widened substantially to some premium peaked at 12 times its prewar 50% toward the end of the 19th cen- level, and it was only stabilized, in par- tury. Following the change in 1892 allel to a significant deflation, when the from the silver florin to the gold crown currency reform of the newly-founded as a new currency, the gold parity was OeNB was implemented. maintained until WW I (chart 2). Repeated government interventions Attempts to find simple associations to increase the money supply were mir- over the period from 1970 to 2015 be- rored in the erratic movement of the tween the current account (as a per- silver premium. After the paper florin centage of GDP) and effective exchange had been debased from silver in 1848, rate developments have not delivered the premium fluctuated widely from a meaningful results. Taking three-year low value of just 5% in 1856–58 to averages, there is just a very weak nega- some 40% in 1861. The low premium tive correlation (R2=0.04) between the episodes are explained in Pressburger current account as defined above and (1966) by a number of special factors percentage changes in the REER, and including (1) the reduction of banknotes only for the subperiod from 1970 to in circulation (1865 and 1874–1876); 1993. Interestingly, for the subperiod and (2) the preparation of the Vienna from 1994 to 2014, the correlation Currency Convention of 1857 and the turns positive (R2=0.20). It is quite

Chart 2 Silver premium (1848–1899) and gold premium (1863–1914) Index, price of paper money=100 150

140

130

120

110

100

50 1845 1850 1855 1860 1865 1870 1875 1880 1885 1890 1895 1900 1905 1910 1915 Silver premium Gold premium

Source: AMECO.

72 OESTERREICHISCHE NATIONALBANK Two centuries of currency policy in Austria

possible that the signing of the Maas- ing the hard currency policy and the tricht Treaty in 1992 and the reforms period following euro adoption, budget following the European currency crisis deficits have often been anticyclical, of 1993 produced sustainable current which means that earlier war financing account surpluses which were not has been replaced by fights against eco- thwarted by real exchange rate appreci- nomic slack and unemployment. Since ations. However, a more convincing in- the late 1960s, as a rough trend, interpretation would require additional creasing budget deficits as a percentage analytical work. of GDP have been associated with ap- There is also a possible interconnec- preciations of the NEER and vice versa tion between current account develop- (chart 3). Theory is ambiguous about a ments and developments in relative in- positive or negative association between terest rates: Whenever the current ac- the two variables, but it is widely account posts an increasing deficit, a rise cepted that growing budget deficits in the domestic interest rate could, may lead to higher domestic interest ceteris paribus, counter an incipient rates, which entail capital inflows and depreciation. Actual data for the period cause the domestic currency to appre- from the early 1970s to the late 1990s ciate.4 In the case of Austria’s hard cur- reveal rather weak negative correla- rency era with its virtually predeter- tions between relative bond yields (the mined monetary policy, the contrac- difference of secondary market yield tionary impact from NEER apprecia- data of ten-year Austrian government tion was compensated by moving in an bonds and, alternatively, German and expansionary fiscal direction, which U.S. rates) and current account bal- was in part corrected after joining the ances. Only in the following period euro area. Another interpretation may after euro adoption are the substantial run via third variables such as the state current account surpluses unequivo- of the economy: In an ailing domestic cally accompanied by comparatively economy automatic stabilizers rise and low interest rates in Austria. tax income suffers, thereby increasing In the post-WW II era, the infla- the budget deficit, but this situation is tionary phases of the 1970s and early also associated with dampened import 1980s also produced comparatively demand, which improves the current high NEER growth rates, while since account and entails a currency appreci- the mid-1990s moderate inflation has ation. However, there is also the well- moved hand in hand with just small known “twin deficits” hypothesis, ac- changes in the NEER. cording to which rising budget deficits will go hand in hand with rising cur- 4.3 Public finances, external rent account deficits and thus lead to accounts and the real sector currency depreciation (Feldstein, 1986). Whenever the printing press is in- As mentioned earlier, for a long volved, there is also an obvious connec- time real sector objectives, such as tion between the financing of govern- more GDP growth and lower unem- ment budgets and inflation. In Austria ployment rates, played at best an indi- this is particularly true for the early rect role in monetary and exchange rate 19th century. In recent decades, cover- policies. This certainly holds for the

4 For an overview of theoretical arguments and empirical findings on the relationship between budget deficits and exchange rates, see Hakkio (1996) and Saleh (2003).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 73 Two centuries of currency policy in Austria

Chart 3 Nominal eﬀective exchange rate and budget balance from 1961 to 2015 (with polynomial trends of the 4th order) % 8

–2

–4

–6

–8 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015 NEER (percentage change) Budget balance (% of GDP) Polynominal trend of NEER (percentage change) Polynominal trend of REER (percentage change)

Source: OeNB, AMECO and author’s calculations.

19th century and for the early 20th cen- from wars and war-related debt financ- tury up to and including WW I. Inter- ing. During the 19th century the financ- dependencies between exchange rates ing needs of belligerent countries were and the real sector became relevant in met primarily by printing money, liter- the interwar period with the outbreak ally by banknotes and state notes and by of the Great Depression. In the extending central bank credit to gov- post-WW II era the dominant policy ernments. In the aftermath of wars, intention was to adjust real sector vari- currency reforms often became neces- ables, in particular productivity, in sary to allow for a new start (as in 1811, such a way as to steer the targeted 1859, 1866, 1925 and 1947). average market exchange rate close to In the years following WW I, ex- a perceived equilibrium exchange rate. change rate policy proper evolved as an active instrument of steering the mac- 5 Summary roeconomy with the main objectives of This paper investigates Austria’s cur- controlling inflation and fostering pro- rency policies over the last 200 years, ductivity advances. An early appear- attempting to sketch historical develop- ance of an anti-inflationary exchange ments and uncover regularities and in- rate policy came a few years before terconnections with other economic WW I when the OeNB was mandated variables. During much of the first to keep the exchange rate of the crown 100 years under observation, Austria’s stable against other major currencies. exchange rate policy was targeted at es- During most of the period consid- tablishing and maintaining the credibil- ered here, Austrian currencies were ity of technical currency parities, in subject to fixed exchange rates, in the spite of activities like coinage debase- form of silver and gold standards in the ment and excessive printing of paper 19th century, as a gold-exchange stan- money. Dramatic changes in the mar- dard in much of the 20th century, and ket values of currencies mostly resulted with the euro as the single currency in

74 OESTERREICHISCHE NATIONALBANK Two centuries of currency policy in Austria

the early 21st century. In between, brief side effects. Given Austria’s adoption of phases of exchange rate flexibility pre- the euro, national exchange rate policy vailed, such as for the paper florin in has been relinquished in favor of a com- the 1880s and early 1890s and the mon currency which itself is floating post-Bretton Woods era of the early vis-à-vis third currencies. Austria’s 1970s. Before long, however, they transformation from one of Europe’s would give way to another peg system. few great powers before WW I to a A prominent form of linking the ex- small open economy closely tied to change rate to a stability anchor was the neighboring Germany thereafter largely “hard currency policy” of the 1980s explains the change from the early and 1990s, which served Austria well, specie standards to a preference for sta- allowing it to achieve the goals men- ble rather than flexible exchange rate tioned above without apparent negative regimes.

References Butschek, F. 1985. Die österreichische Wirtschaft im 20. Jahrhundert. Vienna (Böhlau Verlag). Chaloupek, G. 2003. Carl Menger’s contributions to the Austrian currency reform debate (1892) and his theory of money. Paper presented to the 7th ESHET Conference. Paris. 30 Janu- ary to 1 February. Feldstein, M. 1986. The budget deficit and the dollar. NBER Macroeconomics Annual 1986. Cambridge, MA. Flandreau, M. and J. Komlos. 2001. How to run a target zone? Age old lessons from an Austro-Hungarian experiment. CESifo Working Paper 556. September. Hakkio, C. S. 1996. The effects of budget deficit reduction on the exchange rate. Federal Reserve Bank of Kansas City. In: Economic Review. Third quarter. 21–38. Handler, H. 1989. Grundlagen der österreichischen Hartwährungspolitik: Geldwertstabili sierung, Phillipskurve, Unsicherheit. Vienna (Manz Verlag). Handler, H. 2007. Vom EWS zum Euro: DM und Schilling Hand in Hand. In: Gehler, M. and I. Böhler (eds.). Verschiedene europäische Wege im Vergleich: Österreich und die Bundes- republik Deutschland 1945/49 bis zur Gegenwart. Innsbruck–Vienna–Bolzano (Studien Verlag). 285–304. Jobst, C. and H. Kernbauer. 2016. The quest for stable money: Central banking in Austria, 1816–2016. Campus. Frankfurt. New York. Kernbauer, J. 1991. Währungspolitik in der Zwischenkriegszeit: Geschichte der Oester reichischen Nationalbank 1923–1938. Oesterreichische Nationalbank. Vienna. Klausinger, H. 2006. Oskar Morgenstern als wirtschaftspolitischer Berater in den 1930er-Jahren. Vienna University of Economics and Business. Department of Economics Working Paper Series 98. März, E. 1990. Die große Depression in Österreich 1930–1933. Wirtschaft und Gesellschaft 16(3). 409– 437. Nauschnigg, F. 2003. Kapitalverkehrsliberalisierung – Die österreichischen Erfahrungen. Wirtschaft und Gesellschaft 29(1). 63–72. Pressburger, S. 1966. Oesterreichische Notenbank 1816–1966. Geschichte des Oester reichischen Noteninstituts. Oesterreichische Nationalbank. Vienna. Saleh, A. S. 2003. The budget deficit and economic performance: A survey. Working Paper 03–12. Department of Economics. University of Wollongong. September.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 75 Two centuries of currency policy in Austria

Schmitz, S. W. 2016. The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator. In this volume. Yeager, L. B. 1998. From gold to the euro: The international monetary system in retrospect. In: Dowd, K. and R. H. Timberlake, jr. (eds.). Money and the nation state: The financial revolution, government, and the world monetary system. The Independent Institute. Oakland. Transaction Publishers. 77–104.

76 OESTERREICHISCHE NATIONALBANK The financial relations between the Nationalbank and the government1

The privilegirte oesterreichische National-Bank was founded in 1816, but the central govern- Doris Prammer, ment has held a major share in the Austrian central bank only since 1955. However, the Lukas Reiss, government has always participated significantly in the Nationalbank’s profits via various Walpurga channels. In this paper we analyze the financial relations between the Nationalbank and the Köhler-Töglhofer1 Austrian (or Austro-Hungarian) government over the last 200 years. While today, the Austrian government’s share in the effective profit distribution of the Oesterreichische Nationalbank (OeNB) is 100% (and thus much above 19th-century levels), the OeNB’s holdings of Austrian government debt were relatively small from the mid-1950s until recently. The Eurosystem’s Public Sector Purchase Programme is currently leading to a significant increase in holdings of Austrian government debt recorded at OeNB’s balance sheet. It will likely go somewhat above the levels observed from the 1830s to the 1890s, when average inflation was around ½%. JEL classification: N13, N14, N43, N44 Keywords: economic history, public debt, inflation, central banks

The currency reform of 1816 involved the link between this relationship and the transfer of the right to issue the periods of very high inflation banknotes to a fully privately owned around the two World Wars. Section 4 bank, the privilegirte oesterreichische concludes. National-Bank. The Nationalbank formally remained in (majority) private 1 The relationship between ownership until the Nationalbank Act central banks and governments was passed in 1955. It was fully nation- in general alized only in 2010. However, in com- While formally independent, central pensation for granting the Nationalbank banks in most western economies are the privilege to issue banknotes, the now owned by the government. Even if Austrian government typically had a say central banks are privately owned, the in the appointment of central bank way they function differs from that of managers and has always participated in other private sector entities as eco- the Nationalbank’s profits. nomic ownership may differ from for- In the following article, we will – mal ownership. The most important based on Jobst and Kernbauer (2016) – rights of the economic owners of a discuss the development of the relations typical private corporation are the right mentioned above between the Nation- to participate in profits and the right to albank and the government: Section 1 exercise influence on the appointment sketches the relationship between of managers (typically via the non central banks and governments in executive and/or supervisory board). general. Section 2 focuses on the rela- Both the appointment of managers tions between the Nationalbank and and profit participation may be regu- the Austrian (or Austro-Hungarian) lated in special laws, like the current government, in particular its financial Nationalbank Act 1984 (Federal Law relations. Section 3 briefly discusses Gazette I No 50/1984) in Austria or

1 Oesterreichische Nationalbank, Economic Analysis Division, [email protected], [email protected], [email protected]. The views expressed in this paper are exclusively those of the authors and do not necessarily reflect those of the OeNB or the Eurosystem. The authors would like to thank Clemens Jobst for Refereed by: providing both data and valuable comments. Furthermore, we would like to thank OeNB Governor Ewald Nowotny, Matthias Morys, Doris Ritzberger-Grünwald, Heinz Handler, Hans Kernbauer and Katharina Steiner for helpful suggestions. University of York

MONETARY POLICY & THE ECONOMY Q3– Q4/16 77 The financial relations between the Nationalbank and the government

Box 1

Examples of partly privately owned central banks in the euro area

National Bank of Belgium1 (Nationale Bank van België / Banque nationale de Belgique) The Belgian government owns 50% of shares in the National Bank of Belgium (NBB); the remaining 50% of shares are privately owned. The NBB is directed by the Governor, who presides over the Board of Directors and the Council of Regency2. The Governor and Directors are appointed by the Belgian King upon proposal by the Council of Regency. The NBB’s share capital amounts to EUR 10 million. The Belgian government holds 50% of the 400,000 shares; the remainder is listed on the Euronext Brussels stock exchange. The NBB’s profit is distributed to the shareholders as a first dividend that amounts to 6% of the share capital and hence is independent of the NBB’s profits, and a second dividend, which is distributed after the allocation to the reserve fund. The Council of Regency has set the second dividend at 50% of the net proceeds from the portfolio the NBB has held as a counterpart to its reserves (statutory portfolio) since 2009. The remainder is allocated to the government; it usually amounts to approximately 40% of the distributed profit. In 2015 around 81.5% of the NBB’s overall disbursed profits were distributed to the state; if corporate income tax is taken into account, this share rises to 95%. The NBB’s profit is subject to corporate income taxation, except for the part that is, in the last step, allocated to the government. While being inherently volatile, net profits have exceeded the share capital by a factor of 50 to almost 100 over the past few years.

Banca d’Italia3 The Banca d’Italia is a public law institution (Banca d’Italia, 2016b), held by eligible, (mostly) private entities. It is managed by the Governing Board, which is chaired by the Governor. The Governor is appointed by the President of the Republic, upon resolution by the Council of Ministers and after hearing the opinion of Banca d’Italia’s Board of Directors. Banca d’Italia’s share capital amounts to EUR 7,500 million with a unitary nominal value per share, established by law, of EUR 25,000. Shares may be held by Italian banks, (re-)insurance firms, eligible foundations, social security institutions and eligible pension funds. Hence, the Italian government only holds shares indirectly via social security institutions. However, the statute of the Banca d’Italia affirms that the Shareholders’ Assembly has no power of intervention with respect to the performance of Banca d’Italia’s institutional functions. The Board of Directors distributes net profits up to a maximum of 6% of Banca d’Italia’s share capital to shareholders, and up to 20% of net profits to ordinary and special reserves, respectively. The remainder is disbursed to the government. In 2015 approximately 86% of overall disbursed profits were distributed to the government; if corporate income tax is taken into account, this share rises to 90%. Banca d’Italia’s profit is subject to corporate income taxation. Its yearly net profits amounted to up to two-thirds of share capital in the past few years, while showing a high degree of volatility.

1 See National Bank of Belgium (2016). 2 The Council of Regency consists of the NBB’s Board of Directors and ten members representing the Belgian socioeconomic world, who have to be elected by the General Meeting upon proposal by the social partners and the Ministry of Finance. 3 Banca d’Italia (2016a).

78 OESTERREICHISCHE NATIONALBANK The financial relations between the Nationalbank and the government

the Federal Reserve Act in the United Today, the extent and nature of States. While the Oesterreichische central bank independence has to be Nationalbank (OeNB) was fully nation- assessed on the basis of the relevant alized in 2010, other central banks in legal provisions.3 In the EU, the inde- the euro area are still partly privately pendence of the European Central owned.2 Box 1 provides details on the Bank (ECB) and the national central (partly) privately owned Eurosystem banks (NCBs) is laid down in the Treaty central banks of Belgium and Italy, establishing the European Union and in where the government participates the Statute of the ESCB and of the ECB.4 disproportionally in central bank profits. Historically, one major rationale for 2 The relations between the setting a central bank up as a “privately Nationalbank and the held institution” was that the participa- government tion of private shareholders in the ap- Signaling independence from the state pointment of managers should signal a as well as capital needs undoubtedly limited influence of the government in played a role in the foundation of the the bank’s decision-making process and privilegirte oesterreichische National-Bank thus demonstrate the intention to cre- in 1816. In the years leading up to the ate an institution that would prevent establishment of the Nationalbank, the government from intervening in a lasting and ever worsening inflation, way harmful to monetary stability. eventual state bankruptcy and the Furthermore, private shareholders also return of inflation just thereafter had implied that private capital was raised destroyed faith in the currency (Jobst for setting up such an institution. How- and Kernbauer, 2016, p. 32). ever, a central bank’s ownership structure – in particular private ownership 2.1 Formal ownership – is no longer an argument in the dis- On June 1, 1816, the Nationalbank was cussion of, or an indication for, central incorporated as a private stock corpora- bank independence. tion, with an autonomous management

2 Also the Bank of England, founded in 1694, as well as the Banque de France and the De Nederlandsche Bank, which had both been founded only a couple of years before the privilegirte oesterreichische National-Bank, were set up as privately held joint stock corporations. These three institutions were nationalized after WW II. In contrast, the U.S. Federal Reserve and the Swiss National Bank have remained (partly) privately held institutions to this day. 3 For a discussion of modern academic literature on central bank independence, e.g. Siklos (2007), Cukierman (2008) and Fischer (2015). 4 Article 3 of the Treaty on European Union (TEU) lays down the EU’s economic policy objectives, including “balanced economic growth and price stability, a highly competitive social market economy, aiming at full employment and social progress.” Economic and Monetary Union (EMU) is based on an independent stability-oriented monetary policy, whose primary objective is to maintain the price stability of the euro under Article 127(1) of the Treaty on the Functioning of the European Union (TFEU). Articles 127 through 133 TFEU and the Statute of the ESCB and the ECB call for personal, functional, financial and legal independence of the ECB and the NCBs. Article 130 TFEU states: “When exercising the powers and carrying out the tasks and duties conferred upon them by the Treaties and the Statute of the ESCB and of the ECB, neither the ECB, nor a national central bank, nor any member of their decision-making bodies shall seek or take instructions from Community institutions or bodies, offices or agencies, from any government of a Member State or from any other body.” Article 7 of the Statute of the ESCB and the ECB echoes this statement. The ECB is free to set its policy instruments with the aim of achieving its primary objective of price stability as stipulated in the Treaty. To protect the functional independence of the ECB and the NCBs, Article 123 TFEU stipulates that the monetary financing of budget deficits is prohibited. Also Article 125 TFEU – the prohibition to bail out governments – is to be seen in the context of the functional independence of the ESCB. Hence, in the EU the principle of central bank independence has a quasi-constitutional basis (Bini Smaghi, 2007).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 79 The financial relations between the Nationalbank and the government

acting on behalf of its large number of 2.2 Influence on the appointment of shareholders (Jobst and Kernbauer, managers and on Nationalbank 2016, pp. 37, 252). While the regula- policy tions governing the Nationalbank were Throughout the times of the Habsburg to change multiple times over the empire, the Governor of the National- decades to follow (including the era of bank was appointed by the Emperor the Oesterreichisch-ungarische Bank, i.e. the (Jobst and Kernbauer, 2016, p. 270). Austro-Hungarian Bank), formal own- However, private shareholders always ership was 100% private until after had a say in the appointment of the World War I (WW I). Governing Board (or General Council5), In the interwar period, the OeNB with the extent of their influence was still a private corporation, but the depending on the Nationalbank’s appli- Austrian government held a minority cable statute at the time (according to stake (Kernbauer, 1991, p. 75f, 420). different versions of the statute, for In 1955, the Austrian government be- instance, other Board members were came a 50% shareholder and the re- appointed either directly by the share- maining 50% of shares were allotted to holders or by the Emperor based on selected Austrian financial institutions suggestions by the shareholders; Jobst as well as to social partners like the and Kernbauer, 2016, pp. 37, 40, 42). chamber of commerce and the Austrian As the Nationalbank’s strategy was Trade Union Federation (Nationalbank partly determined by its respective Act 1955). The allocation of shares be- statutes, the government, which set the tween the state and the private sector statutes, always had some influence on remained unchanged until 2006 (Jobst the Nationalbank’s policy. The extent and Kernbauer, 2016, p. 252). of government influence varied over After the Austrian government had time; it was expanded, for instance, in to step in to support the Austrian bank the charter of 1841 (Jobst and Kern- BAWAG P.S.K. in 2006, it was entitled bauer, 2016, p. 42) and reduced in the to acquire the shares BAWAG P.S.K. 1863 charter (Jobst and Kernbauer, held in the OeNB as well as those held 2016, p. 78). by BAWAG P.S.K.’s majority owner, In 1878, the Nationalbank was the Austrian Trade Union Federation. transformed into the Austro-Hungarian After the transfer of these share Bank. As the Austrian and Hungarian holdings, the government held roughly governments were rarely able to agree, 70% of shares in the OeNB (Jobst and the Nationalbank’s independence grew, Kernbauer, 2016, p. 253). In 2010, the in fact, during the time of the government bought the remaining Austro-Hungarian Empire (Jobst and shares that were still privately held. Kernbauer, 2016, p. 113f). This was to This change in ownership was institu- change dramatically during WW I (and tionalized by law in 2011 (Federal Law in its aftermath), when monetary Gazette I No 50/2011), when the federal financing was sizeable (section 2.3). government (Bund) was defined as the After WW I, the Oesterreichische only (possible) shareholder of the Nationalbank (OeNB) was (re)founded OeNB. in 1922 as a (private) stock company

5 The charter of 1878 (establishment of the Austro-Hungarian Bank) introduced a General Council, which had authority over the Austrian and the Hungarian directorate (Jobst and Kernbauer, 2016, p. 120).

80 OESTERREICHISCHE NATIONALBANK The financial relations between the Nationalbank and the government

that became fully operational in 1923. were appointed by the Austrian gov- According to its statutes, which had to ernment or the Federal President; the respect the requirements of the League six remaining members were selected of Nations, it was fully independent from by the shareholders. Important mone- the government (Kernbauer, 1991, p. 57). tary policy decisions (e.g. exchange rate Its President was appointed by the policies) were typically made in con- Federal President (upon proposal by junction with the Ministry of Finance the government), while the other thir- (Jobst and Kernbauer, 2016, p. 226). teen members of the General Council When the OeNB joined the were nominated by the OeNB’s share- Eurosystem in 1999, this framework holders (Jobst and Kernbauer, 2016, changed fundamentally as monetary p. 160, and Kernbauer, 1991, p. 76f). policy decisions have since been taken However, as foreign creditors would by the Governing Council of the ECB. have preferred a foreign President, the The OeNB is represented on the Gov- OeNB was asked to install the position erning Council by its Governor (whose of a Foreign Advisor (Kernbauer, 1991, position has been strengthened, Jobst p. 77). This advisor had the rank of a and Kernbauer, 2016, p. 250ff), who is co-president and his agreement was re- appointed by the Federal President quired for any measures related to cen- upon proposal by the government tral bank policy (Jobst and Kernbauer, (based on a shortlist prepared by the 2016, p. 161). General Council). Right after the annexation of Austria to the German Reich in March 2.3 Government debt holdings were consistently high throughout the 1938, the OeNB was liquidated. The th German Reichsbank replaced the 19 century and peaked after OeNB as the new monetary authority. the World Wars Henceforth, the OeNB’s Governing Participation of the government in the Board had to act on behalf of the Nationalbank’s profits7 has not always German Reichsbank (Jobst and Kern- been in the form of dividends and bauer, 2016 p. 189ff).6 After the end of corporate income tax revenue. Until WW II, when Austria was occupied by 1878 the government did not directly the Allied Forces between 1945 and participate in the official profits of the 1955, the reinstated OeNB had a provi- Nationalbank at all. However, the gov- sional statutory framework based on its ernment was effectively compensated pre-1938 charter (Jobst and Kernbauer, for the privileges given to the privilegirte 2016, p. 252), but all General Council oesterreichische National-Bank via loans members were appointed by the at concessionary rates. Furthermore, government. the government from time to time The 1955 Nationalbank Act estab- issued banknotes itself (so-called gov- lished the state as a 50% shareholder. ernment paper money, see below). Eight of the fourteen members of the These were typically used as a legal General Council (including the tender in parallel to the Nationalbank’s President and the two Vice Presidents) banknotes, and could be exchanged at

6 More than two dozen OeNB employees were immediately dismissed for political or “racial” reasons, others were forced to retire or demoted (Rathkolb and Venus, 2013). 7 The bulk of the OeNB’s profits have always been gained from seigniorage income. Insofar as the OeNB has undertaken commercial activities, earnings from these activities have contributed to profit generation.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 81 The financial relations between the Nationalbank and the government

Chart 1 Domestic public debt held by the Nationalbank % of GDP Revolution Austro- Austro- WW I and Credit- WW II 1999: EMU 100 of 1848 Sardinian Prussian aftermath crisis aftermath common War of War of of 1931 monetary 90 1859 1866 policy

0 1830 1840 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 Banknotes issued by central government Public debt explicitly held by the Nationalbank (from 1999: Eurosystem holdings of Austrian public debt recorded on the OeNB’s balance sheet) Central government debt (1867–1918: including Austria and Hungary) Banknotes (central bank balance sheet+state)

Source: OeNB, ECB, BMF, Jobst and Scheiber (2014), Jobst and Kernbauer (2016).

the Nationalbank. These issuances of Furthermore, the Nationalbank also government paper money are economi- had to convert old government paper cally similar to interest-free loans from money that had been issued before the central bank to the government. the currency reform (which declined Loans to the government were often significantly in the early 1820s; see (temporarily) expanded in times of high chart 3.4 in Jobst and Kernbauer, government financing needs, which 2016). were often caused by the participation During the revolution of 1848, in military conflicts. Chart 1 shows the direct loans by the Nationalbank to the development of the Nationalbank’s state increased (Jobst and Kernbauer, effective holdings of domestic govern- 2016, p. 72) and the government again ment debt (in relation to GDP) since issued government paper money (Jobst 18308. Chart A1 in the annex compares and Kernbauer, 2016, p. 71). Upon these effective debt holdings to the size decision of the government, this gov- of the Nationalbank’s balance sheet total. ernment paper money was redeemed From 1816 to 1847, the average by the Nationalbank in 1854, which in- interest on these concessionary loans creased the government debt explicitly from the Nationalbank was about 2 to held by Nationalbank (Jobst and Kern- 3 percentage points lower than the bauer, 2016, p. 74f and chart 4.2). Loans market yield on government bonds. granted to the government went up again

8 Charts 1 and 4 only start in 1830 as GDP data were not available for the time before that. Moreover, the terms “central government debt” (charts 1 and 5) and “central government budget deficit” (chart 4) refer to administrative data, which are subject to changes in definitions and should be therefore treated with some caution.

82 OESTERREICHISCHE NATIONALBANK The financial relations between the Nationalbank and the government

in 1859 in the course of the Austrian war chart A2 in the annex). At the end of against Piedmont-Sardinia (Jobst and WW I, about one-third of overall gov- Kernbauer, 2016, p. 76). ernment debt was held by the With the 1863 charter, the state Nationalbank, most of the remainder committed to redeem most of its liabil- consisted of long-term war bonds ities against the Nationalbank by the issued at yields of around 6% (Jobst and end of 1866, except for an unremuner- Kernbauer, 2016, p. 145f, and Popovics, ated amount of fl. 80 million at the dis- 1925, p. 81f).10 posal of the state in exchange for the Due to high budget deficits, the monopoly right to issue banknotes. practice of financing the government While this target was reached in 1866, continued after the breakdown of again government paper money was the Austro-Hungarian Empire. In issued to finance the Austro-Prussian November 1922 the public debt held by War (Jobst and Kernbauer, 2016, p. 80f).9 the OeNB amounted to 2,561 billion When explicit profit sharing was crowns (Jobst and Kernbauer, 2016, p. introduced in 1878, preferential lend- 154). Due to hyperinflation, this was ing to the government was no longer less than 10% of Austrian GDP in 1922 necessary to achieve a participation in (chart 1), but more than 300 times the profits. Now the state obtained a share GDP of the Austrian Republic’s terri- of all profits in excess of a minimum tory in 1913 (chart A2). This ratio to return to shareholders (see section 2.4). contemporaneous nominal GDP de- Nevertheless, government paper money creased in 1923/24, partly because of remained in circulation until the 1890s high inflation. From the introduction (Jobst and Kernbauer, 2016, p. 129), of the Austrian schilling (ATS) in 1925 just before the currency switch from until 1930, lending to the government florins to crowns. So in the very early was low when compared to GDP or to- 20th century effective lending from the tal government debt. Nationalbank to the state was negligible. During the Creditanstalt crisis, the As chart 1 shows, this situation was OeNB directly injected capital into to change dramatically during WW I, this troubled bank in 1931 (Jobst and which led to very high budget deficits in Kernbauer, 2016, p. 177); it also pro- the Austro-Hungarian Empire (chart 4). vided extensive liquidity. In 1932, the Nationalbank loans at concessionary government took over most of the rates (originally 1%, later ½%; Jobst OeNB’s lending to Creditanstalt (Jobst and Kernbauer, 2016, p. 147) made up and Kernbauer, 2016, p. 183). The re- a relatively large share of the additionally sulting loan to the government was created debt. After WW I, these loans granted for a concessionary rate of 3%. stood at about 35 billion crowns. While Afterward, public debt in the OeNB’s – due to high inflation – this was less balance sheet remained roughly constant than one-quarter of estimated 1918 until 1938, when the OeNB was liqui- nominal GDP, it was more than 110% dated. The extensive monetary financ- of the nominal prewar (i.e. 1913) GDP ing conducted by the German Reichs- of the Habsburg monarchy (see also bank during World War II (WW II)

9 The Nationalbank granted advances to the government, which were repaid after the war (Jobst and Kernbauer, 2016, p. 81). 10 These yields were similar to the Nationalbank’s average pre-WW I policy rates (see chart 4).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 83 The financial relations between the Nationalbank and the government

will not be discussed in this article; it panded asset purchase programme is briefly sketched in Jobst and (APP)12 is to fulfill the price stability Kernbauer (2016, p. 191f). mandate enshrined in the Treaty. To After WW II, the liabilities of the achieve the objectives of the ESCB and OeNB were mostly balanced by claims to carry out its tasks, the ECB and the of ATS 12.5 billion against the federal NCBs – according to Article 18 of the treasury, replacing claims against the Statute of the ESCB and the ECB – have German Reichsbank (Jobst and Kern- at their disposal credit operations and bauer, 2016, p. 198). This accounting open market operations, i.e. buying approach led to a peak in domestic and selling (government) bonds. The public debt held by the OeNB, as illus- APP was deemed necessary to con- trated in chart 1 (when compared to front downside risks to inflation and either overall public debt or nominal inflation expectations and to achieve GDP). The 1947 currency reform price stability as policy rates were (Jobst and Kernbauer, 2016, p. 201) already constrained by the zero lower significantly reduced currency in circu- bound and the inflation outlook was lation and these claims against the deteriorating, the PSPP respects the government. Various other factors, es- prohibition on monetary financing as pecially transfers of gold restituted to government bonds are not bought in the OeNB by the government, led to a the primary market, which is explicitly further decline until the mid-1950s forbidden under Article 123 TFEU.13 (Jobst and Kernbauer, 2016, p. 212f and table 9.2). The share of debt to 2.4 The government’s share in the GDP shown in chart 1 was also pushed Nationalbank’s overall profit downward by very high nominal GDP distribution growth (real growth of above 10% and Chart 2 shows the government’s share even higher inflation). Afterward, the in “effective distributions” of the Na- ratio of government debt holdings in tionalbank’s profits since 1820. Effec- the OeNB’s balance sheet to GDP was tive distributions consist not only of relatively small.11 profits officially distributed to the state Currently the Eurosystem’s public (from ordinary dividends and special sector purchase programme (PSPP) is profit-sharing agreements) and to pri- leading to an increase of government vate shareholders (such profits are re- debt holdings in the balance sheets ferred to as “explicit profits” below), of the ECB and the Eurosystem NCBs. but also of the government’s interest As of 31 July 2016, the Eurosystem’s savings as well as of corporate income cumulative net purchases of Austrian taxes paid by the Nationalbank. Inter- public debt securities stood at EUR est savings for the government arise (1) 24.6 billion (ECB, 2016). However, the via loans at concessionary rates and (2) ultimate goal of the Eurosystem’s ex- via the issuance of government paper

11 Note that until 1998 chart 1 does not include government bonds held by the OeNB in its investment portfolio (and purchased in secondary markets at market prices). 12 The Eurosystems expanded asset purchase programme consists of the asset-backed securities purchase programme (ABSPP), the third covered bond purchase programme (CBPP3), the PSPP and the corporate sector purchase programme (CSPP). 13 Moreover, the Eurosystem does not carry out transactions in the secondary market in any way which could be perceived as equivalent to acting on the primary market (Mersch, 2015).

84 OESTERREICHISCHE NATIONALBANK The financial relations between the Nationalbank and the government

Chart 2 Shares in eﬀective distributed Nationalbank proﬁts % % of GDP 100 1.0

90 0.9

80 0.8

70 0.7

60 0.6

50 0.5

40 0.4

30 0.3

20 0.2

10 0.1

0 0.0 1820 1834 1848 1862 1876 1890 1904 1918 1932 1946 1960 1974 1988 2002 Government: interest savings Government: explicit proﬁt share Government: corporate income tax Share of private shareholders Total impact on government (right-hand scale)

Source: OeNB, Jobst and Kernbauer (2016). money, which the Nationalbank has to increase in the private share in dis- retire; the latter is economically equal bursed effective profits over the de- to a zero interest rate loan. Effective cades to follow. In 1846, when state distributions do not include, however, liabilities reached their temporary low, transfers to certain risk provisions, the share of effective profit disbursed to profits retained by the Nationalbank the private sector reached a temporary (for the buildup of reserves) or distri- peak of more than 60%. Due to the butions to nonshareholders such as do- significant increase in government nations or funds for the promotion of liabilities with the Nationalbank in the research (e.g. the National Foundation aftermath of the 1848 and 1849 revolu- for Research, Technology and develop- tions, the private sector’s share in the ment or the OeNB Anniversary Nationalbank’s effective distributed Fund).14 profits decreased considerably afterward. In line with the intention behind While retiring government paper the foundation of the Nationalbank, its money was successful, the overall first years were marked by a strong de- receivables from the treasury in the cline of government paper money and Nationalbank’s books peaked in 1854. only a moderate rise in government Moreover, the average interest rate on lending at preferential interest rates. As Nationalbank receivables decreased to the government did not participate at less than 1% (from more than 2% all in explicit profits until 1878, this re- during the preceding decade), while duction in liabilities led to a significant market interest rates kept rising, hence

14 Note that the total amount of “effective distributed Nationalbank profits” paid out to the government shown in chart 2 also includes payments by the Nationalbank to the government which are recorded as equity withdrawals (and not as revenue) in the national accounts. Furthermore, chart 2 refers to the years in which the National- bank’s profits were generated (e.g. the values for 2015 correspond to dividends and corporate income tax paid by the OeNB based on the OeNB’s 2015 results, which were paid out to the government only in 2016).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 85 The financial relations between the Nationalbank and the government

Table 1 Legal provisions for the distribution of the Nationalbank’s effective profits

Charter 1817, 1841, 1863 1878 1887 1899 1910 1922 1930 1932 1955 1981 1999

Reserves (% of 15% 10% profits)1 State (% of remaining 1/3 90% profits) Shareholders4 (% of varying 5%2 5% 4% 8% 6% up to up to 10% nominal capital) provisions, no 6% Reserves (% of share for the 10% remaining profits)1 state Shareholders4 and state 5% to 7%: 4% to 6%: 8% to 10%: 6% to 7%: (% of nominal capital) shareholders 1, shareholders 1/2, shareholders 1/3, shareholders state 0 state 1/2 state 2/3 1/2, state 1/2 6%–7%: 10% to 7% to 8%: share- 12%: shareholders holders share- 1/3, state 2/3 1/3, state holders 2/3 1/5, state 8% to 10%: 4/5 shareholders 1/4, state 3/4 10% to 12%: shareholders 1/6, state 5/6 Treatment of >7%: shareholders >6%: >7%: >10%: >12%: shareholers 1/7, 1/2 to state, Discre- remaining profits4 1/2, share- share- share- state 6/7 1/2 tion3 state 1/2 holders holders holders discretion3 1/3, state 1/4, state 1/4, state 2/3 3/4 3/4 Memo: Nationalbank 0% Varying minority stake5 50% 50% until shares held by central 2006, government ~70% from 2006, 100% from 2010 Source for profit Jobst and Pressburger (1969–1976) Amendments to laws and charters Nationalbank Act 1955 distribution Kernbauer (2016) concerning the Nationalbank (1922, and 1984, respectively 1930, 1932, 1935) (including amendments)

Source: OeNB. 1 Including pension reserves; percentages mostly refer to maximum amounts (i.e. the reserve allocation was smaller in some years). 2 In case of low profits distributed out of the the reserve fund. 3 Remainder up to decision of General Council (typically allocated to reserves). 4 The distribution rules in place until 1955 are to be read as follows (example for 1899): First, shareholders receive dividends up to 4% of nominal capital. Then, up to 10% of the remaining profits are allocated to reserves. The remaining profits are shared equally between shareholders and the state up to the point when overall dividends to shareholders make up 6% of nominal capital. One-third of the remaining amount goes to shareholders, the rest goes to the state. The process stops early in case profits are too small. 5 Kernbauer (1991, p. 75f, 420) mentions the central government‘s share in the Nationalbank in 1923 and 1938. It was below 15% in both cases.

curbing the private sector profit share The 1878 establishment of the considerably. The 1859 state financing Austro-Hungarian Bank was accompa- needs for the war against Piedmont- nied by a profit-sharing mechanism, Sardinia as well as the issuance of gov- allotting a large part of profits in excess ernment paper money after the of a minimum return to shareholders to Austro-Prussian War of 1866 led to the state (after allocation to reserves). further spikes in the state’s share of dis- These rules were then changed with bursed effective profits. Later, the share every new charter (table 1). of private shareholders in disbursed While the charter of 1878 was effective profits varied due to cyclical effective, shareholders were partly developments. protected from profit fluctuations via a

86 OESTERREICHISCHE NATIONALBANK The financial relations between the Nationalbank and the government

reserve fund, while the government did The Nationalbank Act of 1955 not receive an explicit profit share changed profit distribution substan- every year. This protection scheme was tially. After allocation to pension and abolished with the next charter. The reserve funds, the government was en- state’s share in explicit Nationalbank titled to receive one-third of remaining profits was raised steadily in the next profits. Only after that could the two charters. However, its share in General Council allocate up to 6% effective distributed profits decreased (10% as of 1981) of the nominal share over time to less than 50% as govern- value (which initially stood at ATS 150 ment paper money was retired and million) to shareholders. Of the remain- explicit lending to the government was der, one-half was allocated to the state reduced to a constant amount of fl. 30 and the remainder was allocated ac- million (60 million crowns). cording to the General Council’s deci- During WW I, exponentially in- sion. The OeNB’s overall profits increasing concessionary lending to the creased substantially over time, while government drove up the public sector’s the nominal share capital remained share in effective disbursed National- constant until 1999, when it was in- bank profits to well above 90% for the creased marginally to EUR 12 million first time. (roughly ATS 160 million). Therefore, After WW I, the OeNB was (re) the private sector’s share in overall ef- founded in 1922 as a (private) stock fective distributed profits was already company. For the first time, the state below 5% in 1955, and from the 1970s officially held varying minority stakes on it was roughly zero, even though in the central bank (Kernbauer, 1991). lending to the government was at low To make the initial public offering at- levels during this time span.15 There tractive, the dividend that was solely have been two major reforms that in- disbursed to shareholders was set to 8% creased the share of the OeNB’s profits (Jobst and Kernbauer, 2016, p. 252), which is distributed (and not retained), while it had been no more than 4% or and they both solely benefited the gov- 5% since 1878 (table 1). The amount in ernment: in 1992, the OeNB was made excess of the attribution to reserves and subject to corporate income taxation of this minimum dividend was shared (currently, the corporate income tax between the state and the shareholders. rate is 25%); and in 1999, the share of Thanks to stable state liabilities (which (after-tax) profits distributed to the were subject to concessionary interest government was raised to 90% (all of rates), the private sector’s share in ef- which goes to the government). Since fective distributed profits fluctuated the full nationalization of the OeNB in around 25% until 1931. As in 1932 the 2010, the government’s share in the government took over most of the OeNB’s effective distributions has al- OeNB’s lending to Creditanstalt at pref- ways been at exactly 100%, as there are erential interest rates, its liabilities with no longer any private shareholders. the OeNB skyrocketed, increasing its The black line in chart 2 shows that effective share in distributed profits to overall effective profits distributed to roughly 90%. the government were never above 1% of GDP (and only very rarely above

15 We assume that the OeNB charged market interest rates on public debt on its balance sheet from 1955 onward.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 87 The financial relations between the Nationalbank and the government

Chart 3 CPI inflation in Austria % 150 1921: 205% 140 1922: 2,877% 130 120 110 100 90 80 70 60 50 40 30 20 10 0 –10 1810 1820 1830 1840 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 Annual CPI inflation Five-year average of CPI inflation Source: OeNB, Statistics Austria, Jobst and Kernbauer (2016).

½% of GDP) in the years for which tary financing of the government(s). data are available.16 However, these These were also the two periods when figures do not include the government’s OeNB holdings of government debt gains from the devaluation of public were highest in relation to nominal debt via high inflation (which is GDP (chart 1). However, the underlying discussed in the next section). reasons for these large amounts differed. Chart 4 shows that – due to the 3 The contribution of monetary inability and unwillingness of the and fiscal policy to periods of Austrian and Hungarian governments very high inflation to significantly raise taxes (Jobst and Chart 3 shows that since 1830 there Kernbauer, 2016, p. 145) – budget have been only two episodes where the deficits skyrocketed during WW I five-year averages of CPI inflation were while they had been almost zero in the above 10%, namely the period from preceding decades. Furthermore, mone- WW I to the mid-1920s and the after- tary policy rates stood at prewar levels math of WW II.17 Both these periods in spite of increasing inflationary pres- were characterized by very expansive sures, and short-term market interest budgetary policies (as war expenditure rates were actually well below the lom- was mostly deficit financed) and loose bard rate as the monetary financing of monetary policies, including the mone- the war had generated excess liquidity.18 16 Explicit profit distributions and corporate income taxes are based on operating profits. For the calculation of these operating profits, transfers to risk provisions (e.g. for foreign exchange, interest rate, credit and gold price risks) are already taken into account. As transfers to risk provisions – to be determined by the Governing Council – reduce operating profits, the total volume of “effective distributions” and thus the amount of profits distributed to the government as a percentage of GDP is reduced. Transfers to risk provisions recently have been relatively high compared to the pre-crisis years, amounting to as much as 0.2% of GDP in 2012. However, the percentage share of “effective distributions” to the state and private beneficiaries is, if at all, only marginally affected by these risk provisions (if the profit distribution mechanism is not linear with respect to the distributable amount). 17 There were only two more episodes when this average was above 5%, namely the early 1850s (followed by more than a decade of very low inflation) and the late 1970s and early 1980s (first and second oil shock). 18 At the end of WW I, the share of discount and lending transactions in the central bank’s balance sheet had almost vanished (Jobst and Kernbauer, 2016, p. 147ff and chart 6.1)

88 OESTERREICHISCHE NATIONALBANK The financial relations between the Nationalbank and the government

Chart 4 Budget deﬁcits, monetary policy rates and inﬂation % of GDP % 40 40

35 35

30 30

25 25

20 20

15 15

10 10

5 5

0 0

–5 –5 1830 1840 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 Central government budget deﬁcit (1867–1918: including Austria and Hungary) Five-year average of CPI inﬂation (right-hand scale) Policy interest rate (right-hand scale) Source: OeNB, BMF, Jobst and Scheiber (2014), Jobst and Kernbauer (2016).

Therefore, real interest rates were very Austria increased at least twelvefold far below zero until the mid-1920s. As from 1937 to 1945 (Jobst and Kernbauer, the increase in government lending was 2016, p. 193). This banknote overhang mostly financed by issuing banknotes, led to very high inflation in the direct circulation of paper money increased aftermath of WW II. substantially, too. Budget deficits in the The OeNB’s high claims on the newly established First Austrian Re- Austrian government in the mid-1940s public remained very high after the end replaced claims on the German Reichs- of WW I and monetary financing con- bank and were therefore at least indi- tinued until the reestablishment of the rectly related to the monetary financing OeNB in late 1922. Over this period, of WW II. However, in contrast to the OeNB support to the government did early 1920s, budget deficits were fairly not only come in the form of profit dis- modest in the late 1940s, and the tributions in a broad sense (as defined OeNB’s high claims on the government in section 2.4), but especially via a mas- should not be interpreted as a sign of sive devaluation of the existing debt monetary financing of the Austrian stock. government. Charts 1 and 4 do not show data for These two episodes stand in stark the period of WW II (except for infla- contrast to the current situation, where tion), when Austria was part of the – should the PSPP’s pace be maintained German Reich. German fiscal policy – the Nationalbank’s holdings of Aus- was very expansive in that period and trian government debt should be around supported by the German Reichsbank one-tenth of GDP in early 2017. How- (Jobst and Kernbauer, 2016, p. 191f). ever, fiscal policy has been broadly neu- While CPI inflation was kept very low tral in both, Austria and the euro area during WW II via a price freeze, the around 2015 and is above all focused on amount of banknotes circulating in safeguarding long-term sustainability

MONETARY POLICY & THE ECONOMY Q3– Q4/16 89 The financial relations between the Nationalbank and the government

underpinned by the European Fiscal profit distributions and concessionary Framework while it was highly expan- loans from the central bank, the gov- sive during the World Wars. Further- ernment has always participated signifi- more, the war periods also involved cantly (and disproportionately) in the strong supply restrictions in the pro- central bank’s profits. duction of civilian goods and services, Even while the Nationalbank was putting strong upward pressure on in- fully privately owned, the share of pri- flation and pushing real interest rates vate shareholders in “effective distribu- down. Due to low capacity utilization, tions” of profits generally remained be- relatively high unemployment and (too) low 60%. Since the full nationalization low inflation expectations, inflation in of the OeNB in 2010, the government’s the euro area has recently been signifi- share in the OeNB’s effective distribu- cantly below the target (defined as HICP tions has always been exactly 100%, inflation of below, but close to 2% over and it had been close to 100% even in the medium term), to which the Euro the preceding decades. Nevertheless, system responded with the PSPP and the share of overall effective profits dis- other non-standard policy measures. tributed to the government was never Central bank holdings of domestic above 1% of GDP (and only very rarely government debt are not necessarily above ½% of GDP). These figures, how- the major driver of inflation in peace ever, do not include the government’s times. Chart 1 shows that from the gains from the devaluation of public debt 1830s to the 1890s government debt via high inflation. holdings were far above the levels seen The Nationalbank’s lending to the from the mid-1950s to 2014. Neverthe- government was consistently high until less, average inflation during the former the late 19th century. It peaked during period was at around ½%, while it was WW I and in its aftermath as well as above 3% in the latter (chart 3). Further- directly after WW II. Since the mid- more, high (low) holdings of govern- 1950s, the OeNB’s holdings of Austrian ment debt do not inevitably indicate government debt have been relatively strong (weak) central bank indepen- small. dence (like during and after WW I). Currently the Eurosystem’s public sector purchase programme (PSPP) 4 Conclusions will lead to an increase of the holdings The Austrian government had not held of Austrian public debt recorded on the a significant part of Nationalbank shares OeNBsbalance sheet. However, the ex- until after WW II, and it has been the panded asset purchase programme (APP) OeNB’s sole shareholder only since – and therefore also the PSPP – aim at 2010. However, it has always influ- fulfilling the price stability mandate en- enced the appointment of Nationalbank shrined in the Treaty and were deemed managers and – until the adoption of necessary as policy rates were constrai the euro – had a say in the OeNB’s ned by the effective lower bound and be- business and policy operations. More- cause the inflation outlook had been over, due to special rules applicable to deteriorating.

90 OESTERREICHISCHE NATIONALBANK The financial relations between the Nationalbank and the government

References Banca d’Italia. 2016a. About us. www.bancaditalia.it/chi-siamo/index.html (retrieved on June 1, 2016). Banca d’Italia. 2016b. Bank of Italy Legislation www.bancaditalia.it/chi-siamo/funzioni-governance/disposizioni-generali/index.html (retrieved on June 1, 2016). Bini Smaghi, L. 2007. Central bank independence: from theory to practice. Speech at the Good Governance and Effective Partnership conference. Budapest. Hungarian National Assembly. April 19. Brandt, H.-H. 1978. Der Österreichische Neoabsolutismus. Staatsfinanzen und Politik 1848–1860. Volume 2. Göttingen. Cukierman, A. 2008. Central bank independence and monetary policymaking institutions – Past, pesent and future. In: European Journal of Political Economy 24. 722–735. ECB. 2016. Asset purchase programmes. www.ecb.europa.eu/mopo/implement/omt/html/index. en.html (retrieved on August 11, 2016). Federal Law Gazette I No 50/2011. www.ris.bka.gv.at/Dokument.wxe?Abfrage=BgblAuth&- Dokumentnummer=BGBLA_2011_I_50 (retrieved on June 1, 2016). Federal Law Gazette I No 50/1984 (last amended on July 20, 2016). www.ris.bka.gv.at/Geltende- Fassung.wxe?Abfrage=Bundesnormen&Gesetzesnummer=10004409 (retrieved on July 20, 2016). Fischer, S. 2015. Central Bank Independence. Speech given at the 2015 Herbert Stein Memorial Lecture National Economics Club. Washington, D.C. November 4. www.federalreserve.gov/ newsevents/speech/fischer20151104a.htm (retrieved on August 2, 2016). Jobst, C. and H. Kernbauer. 2016. The Quest for Stable Money. Central Banking in Austria, 1816–2016. Campus. Frankfurt. New York. Jobst, C. and T. Scheiber. 2014. Austria-Hungary: from 1863 to 1914. In: Bank of Greece, Bulgarian National Bank, National Bank of Romania, Oesterreichische Nationalbank (eds.). South-Eastern European Monetary and Economic Statistics from the Nineteenth Century to World War II. Kernbauer, H. 1991. Währungspolitik in der Zwischenkriegszeit. Geschichte der Oester reichischen Nationalbank von 1923 bis 1938. Vienna. Mersch, Y. 2016. Monetary policy in the euro area: scope, principles and limits. Keynote speech at the Natixis Meeting of Chief Economists. Paris. June 23. National Bank of Belgium. 2016. About the National Bank. www.nbb.be/en/about-national-bank (retrieved on June 1, 2016). OeNB. 2016. Historical time series compiled on the occasion of the OeNB’s bicentennial. Volume 4: macroeconomic indicators. Forthcoming. Popovics, A. 1925. Das Geldwesen im Kriege. Vienna. New Haven. Pressburger, S. 1969–1976. Das österreichische Noteninstitut 1816–1966. Part I 1–3 (1703–1878) and Part II 1–4 (1878–1922). Vienna. Rathkolb O. and T. Venus. 2013. Reichsbank offices 1938–1945 in focus: main Reichsbank office Vienna. Vienna. Siklos, P. L. 2008. No single definition of central bank independence is right for all countries. In: European Journal of Political Economy 24. 802–816.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 91 The financial relations between the Nationalbank and the government

Annex

Table A1 Background information on the variables presented in this paper

Variable Territory Comment Interpolated Missing years Source years

Nominal GDP Austrian Empire / Missing years for nominal GDP have been 1830–1869, OeNB (2016) Austro-Hungarian Empire interpolated using real GDP growth and 1914–1818 until 1918; Republic of (rescaled) CPI inflation. Data for the period Austria from 1919 before 1870 were retrieved by assuming identical nominal growth in the overall empire and in the current territory of the Republic of Austria. Banknotes in Data refer to banknotes recorded as liabilities of the Nationalbank (plus 1919 Annual circulation banknotes issued by the Austrian government, if applicable). Since 1999, this statements of figure has corresponded to the OeNB’s share in euro banknotes in circulation. the Nationalbank The figure for 1919 was not used as it is distorted due to the then ongoing transition from Empire to Republic. Public debt held by Austrian Empire / Until 1998, this figure excludes government 1919–1920, Annual Nationalbank Austro-Hungarian Empire bonds held in the investment portfolio. From 1938–1945, statements of until 1918; Republic of 1999, this figure may include small amounts of 1952 the National- Austria from 1919 (i.e. general government debt which is not part of bank, ECB excludes public debt (core) central government debt. instruments of foreign countries) Balance sheet size of Where necessary, the size of the Nationalbank‘s balance sheet has been 1841, 1848, Annual Nationalbank adjusted for the amount of government paper money that was convertible at 1938–1944, statements of the Nationalbank. 1952 the Nationalbank Policy interest rate 1830–1998: Nationalbank‘s lombard rate. 1999–2008: Eurosystem main OeNB, ECB refinancing rate. 2009–2015: Deposit facility rate. Distributed profits For computation, see section 2.3. 1910–1913, See table 1 1918–1922, 1938–1955 Long-term yield on Austrian Empire / Data available for all years used in our calculations of profit distribution (i.e. OeNB government debt Austro-Hungarian Empire years in which the Nationalbank granted loans to the government at until 1918; Republic of concessionary rates) except for the period from 1915 to 1937, for which we Austria from 1919 used the lombard rate instead. Central government Austrian Empire / Annual deficits for the period from 1914 to 1914–1922 1860–1862, Brandt (1978), budget deficit Austro-Hungarian Empire 1922 were computed assuming that in given 1919–1920, Tafeln zur until 1918; Republic of accounting periods the nominal deficit was 1938–1945 Statistik der Austria from 1919. Data for identical each month. Data are from administra- Österreichischen the period from 1867 to tive accounts and do not follow the national Monarchie, 1918 include data for the accounts definitions. OeNB (2016), central government of the Federal Ministry Central government Empire as a whole and for Public debt includes banknotes issued by the 1914–1918, 1919–1922, of Finance debt the state governments of goverment and explicit debt to the OeNB. 1913–1927 1938–1945 Austria and Hungary. Missing years for public debt from 1914 to 1918 1931, were interpolated using budget deficits as 1934–1937 proxies for the change in debt. Data are from administrative accounts and do not follow the national accounts definitions. Inflation Republic of Austria CPI OeNB (2016)

Source: OeNB.

92 OESTERREICHISCHE NATIONALBANK The financial relations between the Nationalbank and the government

Chart A1 Domestic public debt held by the Nationalbank % of balance sheet total adjusted for government paper money 100 1999: EMU  common 90 monetary policy 80

0 1820 1830 1840 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 Public debt held by the Nationalbank (from 1999: Eurosystem holdings of Austrian public debt recorded on the OeNB’s balance sheet).

Source: OeNB, Jobst and Kernbauer (2016).

Chart A2 Trends in important economic variables during and after WW I and WW II % of prewar GDP (1913 and 1937) 100,000

10,000

1,000

100

1 1913 1915 1917 1919 1921 1923 1937 1945 1947 1949 Budget deﬁcit Public debt explicitly held by OeNB Total banknotes

Source: OeNB, BMF, Jobst and Scheiber (2014), Jobst and Kernbauer (2016).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 93 Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

Clemens Jobst, This article provides an overview of the supply and demand for cash in Austria over the past Helmut Stix1 200 years. After looking at the government, mint and central bank as cash providers and tracing the composition of cash supply, the article presents several stylized facts about the long-run evolution of cash demand. Results show a relatively stable currency-to-GDP ratio, which is remarkable, given the structural changes in the economy and advances in payment technologies over the past 200 years. Among possible explanations is that innovations have significantly increased the use of close cash substitutes like deposits without, however, reducing the currency-to-GDP ratio. Second, factors like the increasing monetization of the economy during the 19th century might have compensated for the cash-saving impact of financial innovations. Moreover, most of the demand for cash might not be susceptible to payment innovations as it is driven by hoarding. Finally, large political disruptions i.e. times of elevated economic uncertainty, and systemic banking crises exert strong and persistent effects on the use of cash. JEL classification: E51, E42, N13, N14 Keywords: currency, cash, long-run development, payment systems, Austria-Hungary, Austria

Providing a means of payment that is the 200 years since the foundation of generally accepted is one of central the Nationalbank.2 This period was banks’ key functions. The issuance of marked by enormous historic changes. banknotes was already the prime objec- Income per capita multiplied, and with tive of the privilegirte oesterreichische it the demand for means of payment. National-Bank when it was founded in With the shift from a predominantly 1816 and has remained a core task ever agricultural to an industrial, and later a since. The actual use of money issued service economy, a bigger share of in- by the central bank is in turn driven to come was earned in markets where a large extent by demand, whose level transactions had to be paid for in not only reflects the quality of the money, thereby further increasing the money issued and the availability of need for a means of payment. At the substitutes, but also the overall evolu- same time, the evolution of a modern tion of the economy and the financial banking system led to the introduction system in particular. Tracking the role of new cash substitutes and made these of cash thus not only provides valuable substitutes available to a wider range of insights into one of the prime “prod- society. The past 200 years thus pro- ucts” offered by central banks, but vide a good perspective for studying serves as a mirror of developments in how these structural shifts and techno- the economy and society as a whole. logical innovations have affected the Against this background, the paper use of cash. Today, some observers be- offers a descriptive overview of the role lieve that electronic money will lead of cash in the Austrian economy over to the complete disappearance of cash

1 Oesterreichische Nationalbank, Economic Analysis Division, [email protected], Economic Studies Division, [email protected]. The views expressed in this paper are exclusively those of the authors and do not necessarily Refereed by: reflect those of the OeNB, or the Eurosystem. The authors would like to thank Edgar Feige as well as the participants in a workshop among the authors of this volume for helpful comments and valuable suggestions. Edgar Feige, 2 We concentrate here on the role of cash in the economy and avoid any discussion of the technical aspects of the Professor Emeritus, production of cash or the role of the Nationalbank in its distribution, processing and authentication. For a history of University of banknote design and production, see Kranister (1989). On the role of the OeNB in cash logistics today, see Schautzer Wisconsin-Madison (2007), and Koch and Schneeberger (2012).

94 OESTERREICHISCHE NATIONALBANK Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

sooner rather than later.3 At the same confidence in financial institutions, time, the substantial per capita hold- however, cash can serve as a store of ings of cash observed in the major eco value. Taken together, these character- nomies are hard to reconcile with the istics help explain not only the historical use of cash purely for domestic transac- role, but also the continuing impor- tions, a phenomenon termed “cash tance of cash as a means of payment and enigma” (Feige, 2012). Over the last store of value. decade, we have even seen the weight This paper is structured as follows: of cash relative to nominal income Section 1 looks at the supply of cash. actually increasing rather than decreas- Here it is important to also consider ing in many economies. coins, which for a long time constituted Most (macroeconomic) historical an important part of currency in circu- accounts of money concentrate on lation. Section 2 then provides an esti- larger aggregates, notably M1 or M3, mate of total currency in circulation and discuss cash only as a subcompo- and compares these figures with the nent of M1.4 This paper focuses on volume of transactions in the economy. cash.5 Cash is of particular interest for The major finding is a surprising stabil- central banks as the monopoly they ity in the relative importance of cash over command over its creation is their key the past 200 years. Section 3 then looks at lever for influencing broader economic some conjectures to explain this stability. developments. But cash has also several Section 4 provides the conclusion. additional characteristics that distinguish it from other forms of narrow 1 Who supplies cash in Austria? money represented in M1: its legal ten- The right to issue currency has long der status, the absence of interest and since been the prerogative of the sover- its accessibility, together with the lim- eign. During the 19th and 20th centuries ited technological requirements for in Austria, the sovereign exercised this storage and transaction when compared right both directly, notably through the with a bank account, for instance.6 minting of coins but also occasionally Moreover, cash payments are fast, through the issuance of government pa- allow payments to be settled immedi- per money, and indirectly by delegating ately, can be used in person-to-person the right of issue to a dedicated institu- transfers and provide anonymity. The tion.7 After 1816 this institution was use of cash for tax evasion and illegal the Nationalbank. Today the issuance activities has led to calls to phase out of banknotes is administered at the large-denomination banknotes or to level of the euro area, while coinage re- eliminate cash altogether (Rogoff, mains a national domain, albeit under 2014). In times of uncertainty and low common rules.

3 See e.g. the discussion in Wolman (2013) or The Economist (2007). 4 See e.g. Friedman and Schwartz (1963) on the U.S.A. or Capie and Webber (1985) on the U.K. For Austria, the historical evolution of monetary aggregates has been analyzed by Komlos (1983, 1987) and Zipser (1997). 5 For the U.S.A., a chapter in Cagan (1965) provides some hypotheses on the relative demand for cash and cash substitutes. 6 The zero bound on the nominal interest rate that is created by the existence of cash has been seen as an impedi- ment in the current economic situation (e.g. Argawal and Kimball, 2015; Rogoff, 2014; Beer et al., 2016). See also Financial Times (2015). The present article does not discuss the pros and cons of eliminating cash. 7 For more information on the resulting seignorage earnings and the mechanisms for their distribution, see Prammer et al. (2016).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 95 Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

1.1 Coins the national central bank puts them Since the Middle Ages, the right to into circulation. In Austria, the issuing mint and issue coins in the Holy Roman body is the mint itself, which also holds Empire – a traditional prerogative of reserves to cover the costs of a possible the Emperor – had in practice been withdrawal of coins should public de- exercised by the territorial rulers; in mand for coins ever decrease. In case the Habsburg lands, thus, under the the mint’s reserves would not suffice to authority of the Habsburg rulers. While withdraw the surplus coins, there is an in the 17th century a large number of additional government guarantee.8 mints still existed in the different terri- tories under Habsburg reign, minting 1.2 Paper money issued by the was increasingly concentrated in Vienna government as well as Kremnica during the 18th The first paper money was put into century. In the 19th century, Vienna circulation in Austria in 1762 by the already accounted for the bulk of coins Wiener Stadtbanco, a government minted in the Austro-Hungarian Em- finance vehicle run under the auspices pire, with production also taking place of the City of Vienna.9 The first issu- in Prague, Alba Iulia and Kremnica, as ance of paper money was meant to tem- well as Milan and Venice (Koch and porarily help finance Austria’s partici- Jungwirth, 1989). After 1918, the only pation in the Seven Years’ War (1756– mint in the First Austrian Republic was 1763) and was due to be withdrawn located in Vienna. The mint was run as a after the end of the war. However, the division of the Ministry of Finance up new paper money, use of which was until 1989, when it was spun off as a voluntary, satisfied a general demand joint stock company and became a and at times was even quoted above its wholly owned subsidiary of the Oester- nominal value in circulating silver coin. reichische Nationalbank (OeNB) (Ertl, As a result, the government decided to 1989). make paper money a permanent feature The introduction of euro cash in of the Austrian monetary system. The 2002 led to significant changes in the amount of paper money in circulation role of the Austrian mint. Today, de- increased rapidly during the Napoleonic nominations and technical specifica- wars. Rampant inflation ensued. tions of coin production are laid down One of the key objectives of the by the Council of the European Union, financial legislation introduced in 1816 and the ECB approves the volume and was the withdrawal of the depreciated value of coins to be issued each year. paper money and its replacement by While banknote issuance within the new florin (fl.) notes convertible into euro area is centralized at ECB level silver. The issuance of the new notes (see below), it is the individual euro was to be entrusted to the newly area countries that continue to be re- founded independent central bank. sponsible for the issuance of coins From then on, the Nationalbank had a (ECB, 2007). Typically, the issuing monopoly of note issuance, at least in body is the treasury in the national fi- principle. In reality, the notes issued by nance ministries, while the national the Nationalbank and the government mints physically produce the coins and circulated in parallel for long periods of

8 Austrian Coinage Act 1988. 9 For details on the following historical account, see Jobst and Kernbauer (2016).

96 OESTERREICHISCHE NATIONALBANK Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

Chart 1 Composition of currency in circulation Fl. million (after 1900: values in crowns converted into ﬂorin) 1,800

1,600

1,400

1,200

1,000

800

600

400

200

0 1820 1825 1830 1835 1840 1845 1850 1855 1860 1865 1870 1875 1880 1885 1890 1895 1900 1905 1910 Notes issued by the Nationalbank Notes issued by the government Coins (estimates for selected years)

Source: See data annex. the 19th century (chart 1). In the years time the government refrained from immediately after 1816, the old gov- issuing its own notes. The last instance ernment-issued notes were still in cir- of government-issued paper money was culation, but were gradually retired or the short-lived Allied military schillings exchanged for notes of the National- printed by the United States as a bank, so that by the 1830s the major substitute currency for Allied troops part of total paper money in circulation entering Austrian territory in 1944. consisted of Nationalbank notes.10 Due These notes were withdrawn and re- to war and financial pressures, the gov- placed by Austrian schilling (ATS) ernment reverted to issuing its own notes issued by the re-established OeNB notes after 1848, and again after 1866. in December 1945. The notes issued to finance the sup- pression of the 1848 revolution were 1.3 Paper money issued by the relatively quickly retired by being con- Nationalbank verted into Nationalbank notes in 1853, In Austria, the second most important while the government notes issued issuer of paper money after the govern- during the Austro-Prussian War of ment was the Nationalbank. The impe- 1866 remained in circulation for 40 rial decrees founding the Nationalbank years and at times accounted for more in June 1816 and the bank’s first stat- than half of total paper money in circu- utes in 1817 granted the bank the right lation in the Habsburg empire. to issue banknotes that enjoyed the The volume of paper money in- privilege of being accepted in all pay- creased dramatically during the World ments to the state at par with silver War I (WW I) and the early years of coin. In 1841 the Nationalbank became the First Austrian Republic, but this the sole bank to receive that right. This

10 The old notes had no fixed exchange rate versus Nationalbank notes, but their market price was kept stable by the Nationalbank’s continuous conversion operations.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 97 Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

effectively legalized the bank’s de facto subject to a tax of 5%. The new rules monopoly. As a result, and despite allowed for a more flexible manage- voices favoring some version of free ment of note issue, as temporary spikes banking, Austria – unlike many other in demand could be accommodated by countries – never saw the co-existence paying a tax, while the payment of the of several note-issuing banks.11 tax limited incentives for the National- From the very beginning, the bank to overly increase its permanent Nationalbank faced various restrictions issuance. In 1911, the limit was raised regarding its note issue. Under the pegs by 50%. With the onset of WW I, all to silver, and later gold, the de jure and cover requirements were suspended.14 de facto convertibility of banknotes After the post-WW I hyperinflation, into precious metal provided a check to in 1923 the legislator returned to prewar the overissue of notes. Independent of principles with some modifications. As this convertibility, the bank was at before 1914, banknotes in circulation times also subject to statutory limits. (as well as all other sight liabilities) had The first quantitative rule was estab- to be covered by at least 40% in gold lished in 1858, limiting note issuance to and other specified foreign assets. Un- three times the available stock of silver.12 like before 1914, however, coverage Following the English example, the was allowed to drop below 40% subject charter of 1862 established a maximum to a progressive tax on the excess issue.15 fiduciary (i.e. not covered by silver) Cover requirements thus no longer im- issue of fl. 200 million, while the posed any absolute limit on note issu- amount of banknotes surpassing this ance. The note issuance tax was finally threshold had to be fully covered by the abolished in the Nationalbank Act of Nationalbank’s silver (and later also 1955. From then on, the cover require- gold) reserves.13 With the growth of ments no longer acted (primarily) as a the economy, the limit on the fiduciary break on the note issuance on the issue proved overly restrictive and was central bank’s liability side, but as re- replaced by a more flexible limit in strictions on the composition of the 1887, which emulated the rules of the central bank’s assets.16 German Reichsbank. Now fiduciary With the introduction of euro issuance was allowed to exceed fl. 200 banknotes and coins in 2002, the million, with any excess amount issued Eurosystem assumed control over cash

11 Hungarian demands for the establishment of a separate Hungarian note-issuing bank, which followed the conclusion of the Compromise in 1867, were settled by reorganizing the Nationalbank as a dualistic institution, the Austro-Hungarian Bank (Oesterreischisch-ungarische Bank), giving the Austrian and the Hungarian parts of the dual monarchy equal representation in the bank’s decision-making bodies. See Jobst and Kernbauer (2016). 12 Note that cover requirements evolved over the 19th century and were specified only relatively late. Article 14 of the 1816 statutes stipulates simply that “the Bank should never issue more notes than the funds assigned to their conversion would allow.” Article 15 of the 1841 statutes states, “It is incumbent upon the directors of the Bank to set from time to time such a ratio between note issue and specie reserves that complete fulfillment of this duty [to pay the face value of banknotes in legal silver coin on demand is assured.” (authors’ translation; see Pressburger (1959)). 13 The fiduciary issue had to be covered by specified domestic income-generating assets, but this did not impose a constraint on increasing the value of banknotes in circulation. 14 The issuance limit had also been suspended in the wake of the 1873 stock market crash (May 13, 1873, until October 11, 1874). See Jobst and Kernbauer (2016). 15 40% coverage was to apply after the resumption of specie payments, which were never introduced, however. In the meantime, lower but gradually rising percentages were to be applied. See Articles 85–89 Nationalbank Act 1922. 16 Article 62, paragraph 1 Nationalbank Act of 1955.

98 OESTERREICHISCHE NATIONALBANK Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

in circulation. While in practice the holdings of financial institutions and national central banks (NCBs) continue government, even though not always to be responsible for the physical issu- recorded, are unlikely to have been suf- ance and management of cash, the cir- ficiently large and volatile to affect the culation of currency is no longer at- long-term trend in paper money circu- tributed to individual countries within lation. The situation is different for the euro area. Today, the number of coins, where data on production is banknotes in circulation reported on available but estimations of actual cir- the OeNB’s balance sheet refers to a culation are affected by a lack of evi- fixed percentage of the circulation of dence on the initial stock as well as the banknotes in the entire euro area statistical problems created by the fact (Krsnakova and Oberleithner, 2012). that full-bodied coins could be melted down for industrial purposes or im- 2 Secular trends in the demand ported and exported without being for cash properly recorded. Add to that the pos- In the short run, the government and sibility, at least for the earlier years, the central bank, as suppliers of cash, that foreign coins were used in domestic can increase or decrease the amount of transactions, for which no information cash in circulation at will, at least in is available either. These knowledge principle. In the long run, however, the gaps are unfortunate. Evidence for price level adjusts, and currency in cir- other countries indicates that before culation mainly reflects the demand for the 20th century, coins accounted for an cash and the relative attractiveness of important part of total cash in circula- other assets that can be used as a means tion, so the omission of coins creates a of payment or store of value. Over the potentially large bias in any time series last 200 years, both the price level and for currency in circulation. The prob- the volume of transactions potentially lem is aggravated by the fact that esti- to be concluded in cash have increased mates for other countries vary enor- sharply. To understand the role of cash mously, so that national statistics are in the economy it is thus useful to look highly idiosyncratic and generalizations at real cash balances (i.e. adjusted for difficult to make. Around 1910, the inflation) or the ratio of currency in share of coins in total currency in cir- circulation to nominal income (i.e. ad- culation (always outside the central justed for inflation and income growth). bank) ranged from 79% in the U.K. to Before doing so, however, we have to 11% in Canada, while Germany occu- discuss the quality of data on currency pied the middle ground with 62%. In in circulation, notably the limited evi- addition, ratios differed not only be- dence available on coins in circulation tween countries but could also change for the 19th century. significantly over time, as e.g. in the U.S.A., where the share of coins in- 2.1 What role did coins play in creased from 5% immediately after the 19th-century Austria? Civil War to 30% before WW I.17 Paper money in circulation can be rela- As in the U.S.A., the turbulent tively well measured, as the amount monetary history in Austria very likely put into circulation is known and the affected the composition of the country’s

17 Data on the U.K. from Capie and Webber (1985), and on Germany from the Deutsche Bundesbank (1976). For the U.S.A. and Canada, see Metcalf et al. (1998).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 99 Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

currency in circulation. Repeatedly, dearth of divisional coins. In 1868, the heavy issuance of paper money and the government tackled the consequences suspension of the convertibility of paper by sharply reducing the silver content money into silver or gold led to the of the 20, 10 and 5 kreuzer pieces. The emergence of a premium on full-bodied gold standard legislation of 1892, intro- coins and the likely disappearance of ducing the gold crown as the new cur- full-bodied coins from circulation rency unit minted as K 10 and K 20 (chart 2). The effect was potentially pieces, finally established a truly fidu- even more disruptive, as even silver ciary divisional coinage. The K 5 and K coins of relatively small denominations 1 coins were minted in silver, but the had (proportionally) the same fine decline in the market value of silver silver content as the fl. 1 silver coin. made their withdrawal or hoarding Following the Austro-Bavarian mint highly unlikely, and lower denomina- convention of 1753, which formed the tions were minted in low-value nickel basis for the Austrian coinage system and bronze (Rieder, 2011). From 1903 until 1857, all coins down to the onward, the Nationalbank also put gold groschen (= 3 kreuzer or fl. 0.05) con- coins into circulation, which were tained proportionally the same amount never used widely, however, and in fact of fine silver, with only the denomina- were soon largely redeposited with the tions below 3 kreuzer being minted in central bank and the treasury. copper (Rieder, 2011). The high silver Whether silver coins circulated or content meant that even small divi- not thus critically hinged on whether sional coins faced the risk of disappear- the Austrian currency traded at par ing from circulation as soon as silver with silver. With the exception of a few traded at a premium. In order to facili- months in early 1859, this was not the tate trade, Austria and the German case for any of the years between 1848 Customs Union (Deutscher Zollverein) and 1878. On the face of it, the pre- concluded a coinage treaty in 1857, mium on silver, which averaged 16% harmonizing their circulation of silver over the thirty-year period, would im- coins (Willis, 1896). Austria slightly ply that no silver coins circulated. At lowered the silver content of its cur- the same time, there are several known rency and introduced a new florin ö.W. instances where the silver florin was (österreichische Währung, i.e. Austrian used as a unit of account, e.g. in the de- currency) to replace the former florin nomination of bonds or share capital Conventionsmünze (CM). While the treaty and the setting of railroad freight rates. prescribed full silver content only down It is not known whether actual pay- to the fl. 0.25 piece, the smaller de- ments in these transactions were done nomination silver coins down to fl. using actual silver coins or paper 0.05 continued to be minted with a rel- money, by taking the daily premium on atively high silver content and were silver into account. The premium on only about 10% underweight (Rieder, silver thus gives no clear indication of 2011). As a result, when in the wake of whether silver coin circulated or not.18 massive issuance of government paper In 1873, however, the price of silver in money in 1866 the premium on silver international markets began to decline increased sharply, Austria again faced a gradually, so that by 1878 the premium

18 In k. k. Finanzministerium (1892, p. 258), it is argued that after 1848 silver remained initially in circulation and only disappeared with the issuance of large silver-denominated loans in the 1850s.

100 OESTERREICHISCHE NATIONALBANK Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

Chart 2 Legal monetary standards, actual exchange rates and their eﬀects on coins in circulation Index, price of paper money = 100 160

140

120

100

1820 1830 1840 1850 1860 1870 1880 1890 1900 1910

Legal silver standard Legal gold standard

Silver coins circulate at face value Silver coins do not circulate at face value Silver coins circulate at face value

Export of silver coins can be proﬁtable Export of silver coins cannot be proﬁtable

Exchange rate on Augsburg Price of silver Price of gold Source: Wiener Börse (1786–1858, 1858–1861, 1861–1875), Der Tresor (1872–1918), authors’ conversion into percentage of par value. on the silver florin disappeared and In the late 18th century, the circula- soon turned into a discount, i.e. the sil- tion of paper money increased rapidly ver florin coin traded at a premium rel- without initially causing paper florins ative to the commodity value of its sil- to trade at a discount to silver. In 1799, ver content. This meant that silver a premium on silver appeared, which coins could circulate again at face value. implies that by then paper money must The government resumed the minting have crowded out the bulk of coins. of silver coins. No information, how- As paper money issuance in 1799 ever, is available on how many of the amounted to fl. 140 million, the value florins minted before 1878 had re- of coins in circulation must have been mained in circulation (trading at mar- something around fl. 150 million during ket prices) or had returned after 1878 the 1780s, when paper circulation was from hoardings or from abroad. still very low. For 1847, the last year While the hoarding and export of before the monetary turmoil caused by coins that accompanied episodes of the revolution of 1848, estimates based monetary turbulence renders standard on a similar method would yield coins techniques for the estimation of the in circulation worth approximately coin stock unworkable, events that led fl. 100 million to fl. 120 million, which to the sudden disappearance of coins as would amount to around 30% of total well as monetary reforms which within currency in circulation (including bank- a short period of time replaced all cir- notes) of fl. 320 to fl. 340 million.19 culating coins with newly issued money With the florin inconvertible into allow at least some educated guesses for silver during most of the second half of selected benchmark years (chart 1). the 19th century, the share of coins in

19 The fl. 100 million to fl. 120 million range is also in line with the direct estimates in k. k. Finanzministerium (1892, p. 258).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 101 Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

total cash in circulation certainly (admittedly rough) estimates for coins declined, even though it is difficult to in circulation. put an exact number on it. Some indica- Deviations of cash holdings from tions can be gleaned from the changeover the long-run trend coincide with clearly from florins to crowns. For 1892, con- identifiable shocks. Wars in 1848/49, temporary economists estimated that 1859 and 1866 led to temporary in- roughly 10% of total cash consisted of creases in cash per capita, while auster- coins, which then presumably increased ity policies in the early 1860s led to a to reach about 15% to 20% before sharp decline in cash per capita. When WW I, depending on whether gold is the substitution of lower-denomination assumed to have circulated or not.20 notes by new crown coins is taken into Compared to Germany, the U.K. or account, growth during the 1890s and France, Austria-Hungary was thus defi- 1900s is no longer below average, but nitely placed among the high-banknote, back on trend. Not surprisingly, real low-coin countries at the beginning of cash holdings per capita fluctuated the 20th century, but was no outlier widely during WW I and the ensuing compared with the U.S.A. and Canada, hyperinflation, with balances first infor example. This means that paper creasing in 1914 and 1915 and then money in circulation should give a rea- declining when inflation accelerated sonable approximation both for the toward the end of the war. After WW I, amount and the trends in currency in real per capita cash holdings increased circulation, at least after 1850. again as the issuance of banknotes out- Due to rampant inflation during paced inflation, at least at the begin- and after WW I, higher denomination ning of the hyperinflation period. With coins disappeared. After WW I, the the stabilization of the Austrian cur- OeNB also collected and regularly pub- rency in 1923, per capita cash holdings lished data on coins in circulation. In returned to their long-run trend. The 1930, (divisional) coins accounted for monetary overhang after World War II 8% of total currency in circulation. By (WW II) was rapidly eliminated by a 1999, this share had declined to 4%. combination of monetary reforms and the release of inflationary pressures 2.2 Currency in circulation shows no that had built up during wartime clear trend rationing. After dropping well below Chart 3 shows the long-run evolution the long-run trend, real per capita cash of cash per capita starting from the holdings had reverted to trend by the early 19th century. To account for 1960s. changes in price levels, cash holdings The changeover from Austrian schil- are deflated using a consumer price ling to euro cash led to a sharp drop in index. Chart 3 shows that since the currency in circulation. Since Janu- 1820s, real per capita cash holdings ary 1, 2002, Austria has been part of a have increased at a remarkably steady single currency area, which means that rate of about 1.7% per year. The pic- cash circulation can no longer be ob- ture is little affected if we include our served at the national level. Per capita

20 Menger (1892, p. 653) calculated the amount of currency in circulation as up to fl. 936 million for 1891; of this, fl. 834 million were banknotes and state notes in circulation, fl. 50 million silver and gold coins, and fl. 52 million divisional coins.

102 OESTERREICHISCHE NATIONALBANK Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

Chart 3 Real cash per capita 1913 crowns 1,000

100

Trend line = 13.77e0.0169 year

10 1820 1830 1840 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 Paper money, real, per capita Total currency in circulation (including coins) OeNB share in euro circulation Source: See data annex. cash holdings based on the OeNB’s holdings seem to be back to the long- share in total euro area cash circula- run trend. tion, which in turn depends on Austria’s The secular increase in per capita share in euro area population and GDP, cash holdings is essentially due to rising however, show a quick recovery after incomes. When tracking the evolution the changeover. Today, per capita cash of transaction volumes, the underlying

Chart 4 Currency in circulation in relation to nominal GDP % 20

0 1830 1840 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 Paper money in circulation Total currency in circulation (including coins) OeNB share in euro circulation Source: See data annex.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 103 Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

driver of cash demand, standard money caused by the euro cash changeover is demand functions therefore include not clearly visible and so is the recovery only prices, but also real income. How- afterward. Today, the ratio of currency ever, reliable estimates of nominal income in circulation to nominal GDP is back or nominal GDP are only available for at 8%. much more recent periods than popula- For the euro area as a whole, the tion and price series. Chart 4 shows EUR 1,035 billion of currency in circu- currency in circulation as a percentage lation translate into per capita cash of nominal GDP starting in 1830. Note holdings of roughly EUR 3,000. As in that the values for nominal GDP before the case of the U.S. dollar, part of these 1870 are linear interpolations between surprisingly high per capita figures can ten-year benchmark estimates and that be explained by circulation outside the even after 1870, when estimates for monetary area. Estimates for the euro real GDP improve, the deflators used area that are based on banknote ship- are rather crude. Therefore, only the ment data put the share of euro long-run trends in chart 4 can be inter- banknotes circulating outside the euro preted with some confidence. area at around 18% of the total stock. The single most remarkable obser- However, even if this estimate consti- vation from chart 4 is the long-run tutes a lower bound due to flows of stability of the ratio between currency currency outside the banking system, in circulation and nominal GDP, which for example via workers’ remittances for most of the time fluctuated between or tourism, it implies that roughly EUR 8% and 12% without any clear trends 2,500 per capita are circulating domes- and is probably even more stable if esti- tically.21 Domestic circulation in Austria mates for coins are included. Chart 4 seems to be even higher. The OeNB’s thereby confirms the evidence gleaned internal estimates based on information from real cash per capita in chart 3, from its cash logistics activities put which likewise showed remarkable domestic circulation in Austria at EUR stability around a long-term trend. 28.3 billion or EUR 3,300 per capita. Again the big outlier is the immediate The importance of cash therefore seems post-1945 period. In contrast to real undiminished. cash, however, cash as a percentage of nominal GDP remains at historically 3 Stability of cash demand is high levels well into the 1960s. This puzzling in light of fundamental means that the below-trend values of structural changes real cash can be explained by the low The long-run stability in the relation- level of real GDP after WW II. By the ship between currency in circulation mid-1960s, the ratio of currency in and nominal GDP is surprising given circulation to nominal GDP started to the fundamental economic changes in decline, reaching what seems to have the economic and institutional struc- been a new equilibrium level of about ture of the economy over the 200 years 6% in the 1980s. Again, the drop under observation. Section 3 looks at

21 On the international circulation of the euro, see ECB (2015). Feige (2012) puts the share of U.S. dollar banknotes held abroad at between 30% and 37%. Accordingly, of the USD 2,950 per capita in circulation in 2011, between USD 1,850 and USD 2,000 were effectively held inside the U.S.A.

104 OESTERREICHISCHE NATIONALBANK Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

three important aspects that exemplify data on wages were systematically col- these changes, with the aim of gaining lected.22 From the available evidence it a better understanding of the (changing) is not clear exactly which profession role of cash in the economy. and activity they relate to, whether payments in kind were included, and 3.1 Changing use of cash – evidence finally how daily or weekly wages could from its denominational structure be converted into a monthly or an One aspect of banknotes and coins con- annual wage, as indications on the an- cerns their usability for settling average nual length of employment and seasonal transactions. Table 1 is an attempt to patterns in wages and hours worked are relate per capita cash in circulation to missing (Sandgruber, 1982). While an economically meaningful measure these limitations create significant chal- of the transaction volume of the average lenges for studies on the standard of population. living, they seem less important for our Specifically, we collected monthly purposes, as we are mainly interested wages for selected years (table 1, col- in rough orders of magnitude. umn 1) whose computation, however, The monthly wages reported in requires some clarification. For the 19th table 1 were thus calculated as follows. century in particular, reliable evidence First, until 1910 we only have information on money wages is difficult to come by. on daily wages, while wages for 1930 Before 1890, when accident insurance are weekly. These daily or weekly wages statistics became available, very little are converted to monthly wages by

Table 1 What can which banknotes pay for? The different denominations of currency in circulation

Year Currency unit Average Currency in circulation Circulation of banknotes with a face monthly value of 1,000 wage Value of Value of Value of Share of % of % of Per capita Number of banknotes coins in currency in coins in banknotes currency in supply monthly in circulation circulation total value in circulation (value) wages circulation per capita per capita of currency circulation equivalent per capita in that can be to one circulation used to pay banknote in % out an average monthly salary

(1) (2) (3) (4) (5) (6) (7) (8) (9)

1840 Florin Conven- 6 5 ≈4 9 ≈35 23 16 1 158 tionsmünze 1880 Florin ö. W. 16 17 n.a. n.a. n.a. 43 16 3 61 1910 Crown 54 48 12 60 19.6 50 18 9 19 1930 Austrian schilling 310 163 14 177 7.8 91 9 16 3.2 1960 Austrian schilling 2,159 2,657 99 2,756 3.6 100 45 1,263 0.5 1980 Austrian schilling 12,495 10,172 390 10,562 3.7 100 73 7,754 0.1

Source: Wages: see text. Circulation and population: see annex. Denominations: Lucam (1861), Lucam (1876) and OeNB (1818–1938, 1957–2015). Note: Column (1) summarizes estimates of average monthly wages as discussed in the text. Columns (2) to (4) summarize per capita values of banknotes and coins in circulation. Column (5) shows the share of coins in the total value of currency in circulation (data for 1840 report the estimate for 1847). Column (6) shows the value of denominations with which monthly wage bills could be settled as a percentage of the overall value of cash in circulation. Column (7) refers to the circulation of banknotes with a face value of 1,000. Column (8) shows the per capita supply of banknotes with a face value of 1,000. Column (9) expresses the number of monthly wages required to buy a banknote with a face value of 1,000.

22 The most important data sources are the annual statistical handbooks, which contain regionally differentiated day wages. For a comprehensive list of sources, see Cvrcek (2013).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 105 Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

assuming 25 working days per month circulating in denominations that were and 52 working weeks for weekly within potential reach of broader wages. Second, wages for 1840 and strata of the population. Table 1 shows 1880 refer to – presumably agricultu- that the denominational structure of ral – minimum day wages, whereas the banknotes changed substantially over figures for 1910 and 1930 refer to wages time – the respective share was 100% in the metal industry. The figures for in 1960. 1960 and 1980 refer to gross industry The significant share of very high wages. Third, the wages quoted for denomination notes (denominations 1840, 1880 and 1910 refer to wages in exceeding the worth of a monthly salary) Bohemia outside Prague. Bohemia was until 1900 illustrates that cash must selected as providing a sort of middle have been used to settle not only small- ground between high-wage regions in value but also large-value transactions. the west and low-wage regions in the As discussed by Kernbauer (2016), the east of the Habsburg monarchy. In 1840, giro system for larger companies for example, Galician wages were 50% (mainly via noncash transactions at the lower than wages in Vienna. Again, the Austro-Hungarian Bank) started to main purpose is to obtain a rough order grow rapidly around 1890. Around of magnitude. 1900 the postal giro system, mainly To assess the usability of cash for used by companies and/or for larger- “normal” transactions, we analyze the value private transactions, started to denominational structure of banknotes quickly expand and gain importance. over time and relate it to monthly Over time, this financial innovation wages. Column 6 shows the total rendered cash less and less important percentage share of all denominations for large-value transactions. This is small enough to be used to settle a one- exemplified by the fact that the denom- month wage bill.23 For example, in ination of the highest-value banknote 1840 the only banknote available for was not adapted to the enormous in- putting together a monthly wage of fl. crease in both real and nominal income 6 was the fl. 5 banknote, as no smaller and remained 1,000 throughout until denominations were in circulation. The 1989.24 As a result, banknotes with a fl. 5 banknote accounted for 23% of face value of 1,000, out of reach the overall value of banknotes in circu- for ordinary households in 1840 when lation. In 1880, three denominations its value corresponded to about 160 (fl. 10, fl. 5 and fl. 1) were suitable for monthly incomes, became the staple the payment of a monthly wage of fl. banknote by 1980, when ATS 1,000 16. Together, these denominations ac- came to only 10% of an average monthly counted for 43% of total banknotes in salary. Correspondingly, the relative circulation. In other words, in 1840 share of the highest denomination only about one-quarter and in 1910 still banknote increased from 16% in 1840 only one-half of the value of cash was to 73% by 1980 (table 1, column 7).

23 Although this is clearly only hypothetical for the earlier years, as wages were paid out daily, a monthly wage reflects a larger transaction that potentially could have been conducted by an average wage earner. 24 The only exception here is the hyperinflation period during the early 1920s, which temporarily saw the issuance of notes with denominations up to 500,000. With the introduction of the Austrian schilling (ATS) in 1925, the OeNB reverted to 1,000 as the highest denomination. Starting in 1989, the OeNB issued an ATS 5,000 banknote.

106 OESTERREICHISCHE NATIONALBANK Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

Chart 5 Monetary aggregates as a percentage of nominal GDP % 90

0 1870 1875 1880 1885 1890 1895 1900 1905 1910 1915 1920 1925 1930 1935 1940 1945 1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 Paper money M1 M2 M3 Source: See data annex.

3.2 Cash and bank deposits of demand deposits (in order to econo- Although currency in circulation has mize on gold to back paper money after been surprisingly constant relative to the adoption of the gold standard in GDP, its importance has changed rela- 1892) all fell into this period. This tive to other assets. Therefore, we ana- development is driven not only by the lyze the evolution of deposits relative to expansion of financial institutions but cash. also by interest rates.26 Saving banks Chart 5 shows the evolution of mon- paid an average interest rate of between etary aggregates M1 and M3 and of 3.8% and 4% in all years from 1890 to cash as a percentage of GDP.25 It reveals 1912 and inflation rates were rather several noteworthy facts. First, bank low, giving rise to high real interest deposits rose steadily from the 1880s rates.27 On balance, however, Komlos until WW I, while currency in circula- (1979) argues that the important driver tion remained relatively stable. This of deposits was transaction costs. The applies both to M1 and M3. In this increasing number of savings banks and period of institutional change in the credit cooperatives as well as the offer financial sector, cash lost importance of financial services through the dense relative to close substitutes. As discussed network of post offices mobilized the in Komlos (1983), the foundation of a savings of the poorer sections of the government-sponsored postal savings population, which had been excluded system and the strong expansion of the from banking before. Overall, this network of bank branches, the rapid induced people to hold deposits rather spread of the postal giro system, the than cash. government’s encouragement of the use

25 For the interwar period, we use M2 due to the lack of an M3 measure. 26 Komlos (1987) econometrically analyzes demand for M1 in this period. His estimates suggest a structural break caused by financial innovation around 1890. 27 Within these 22 years, inflation was below 1% in 11 years, and between 1% and 2% in 5 years.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 107 Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

Chart 6 Deposit-to-currency ratio: comparison with the U.S.A. and Germany Ratio = deposits/currency in circulation, where deposits = M1–currency in circulation 7

0 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 Austria (-Hungary) U.S.A. U.S.A.-constructed Germany Source: See data annex.

Second, chart 5 illustrates the dra- large component of M3) were 3.5%. matic consequences of both World Wars Given positive inflation rates that for capital formation. In the 1950s, the reached 5% in some years, the incentive M3-to-GDP ratio reached only the to hold demand deposits was low. How- level of the 1880s and it took until the ever, regardless of the compositional ef- 1980s to reach the pre-WW I level. fects between M1 and M3, a substantial Third, somewhat surprisingly, the rapid decline in the importance of cash relative proliferation of transaction accounts to M3 is evident. and of other financial innovations after To further analyze the extent of WW II is not reflected in an increase of substitution between cash and deposits, M1 (relative to GDP), whereas M3 a somewhat different perspective is grew substantially. To some extent this provided by the deposit-to-currency is striking given the spread in account ratio (D-C ratio) which expresses the ownerships, the growth of noncash giro amount of money held in bank accounts payments and the continued transfor- relative to one unit of money held as mation of the settlement of wage bills cash (chart 6). The D-C ratio therefore from cash to noncash transfers that reflects the preferences of the private occurred from 1950 until 1980. How- sector regarding the composition of ever, this period was also marked both their financial asset holdings (Friedman by very low interest rates on demand and Schwartz, 1963). Chart 6 shows (M1) deposits and by a sizeable inter- the D-C ratio for M1 deposits for est-rate spread between demand and Austria(-Hungary), Germany and the time (M3) deposits. From 1950 until the U.S.A.28 1970s, interest rates on transaction From 1875 until 1913, the D-C accounts were only 0.75% (with only a ratio evolved almost parallel in all three few exceptional years with interest economies. This demonstrates that the rates of 1%). By contrast, interest financial development during this time rates on savings accounts (which are a was an international phenomenon. The

28 On the construction of the U.S. series, see data annex.

108 OESTERREICHISCHE NATIONALBANK Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

international technology transfer is 3.3 Structural changes in the described in Komlos (1983, p. 142): economy, payments and “Officials were sent to Germany to hoarding learn more about the technology [giro The previous section described how transfers] … and many of the innova- cash lost importance relative to other tions adopted were, in fact, on the assets but remained in fairly stable de- German model.” The evolution of this mand relative to GDP. How can these ratio diverged, however after WW I observations be reconciled? (no data available for Germany). After We suggest two possible explana- WW II, the D-C ratio remained sub- tions. First, the observed stability in the stantially lower in Austria and Germany face of significant structural changes in than in the U.S.A.29 the economy could be due to their The D-C ratio is an indicator of countervailing effects on cash demand. the confidence that the public extends Second, the use of cash might be largely toward banks. In 1931, with the bank- driven by motives which are not parti ing panic in the U.S.A. and the break- cularly susceptible to payment innova- down of Creditanstalt in Austria, the tions. D-C ratio declined sharply. Also, both To explore the first hypothesis, we World Wars led to a considerable compare a structural change in the decline of the D-C ratio. It is interest- composition of GDP with a counter- ing to observe that after major shocks, vailing structural change in payment the D-C ratio tends to remain per- technologies: sistently lower for many years. This is • Although quantitative evidence is most evident in post-war periods. scarce, it can be presumed that ex- Moreover, the Great Depression had a tensive parts of the early 19th-century lasting effect as well. Even after the im- economy still operated using rela- position of a deposit insurance system tively little cash. It is likely that cash in 1932 in the U.S.A., the D-C ratio became more important, in particular continued to be lower than before the in agriculture, where the abolition of crisis. This observation fits well with the last remnants of a feudal system the literature on households’ financial in 1848 should have led to an increase behavior after a financial crisis experi- in wage labor. Moreover, an overall ence (Brown and Stix 2015; Osili and shift out of agriculture into industry Paulson, 2014), which reports per- and services should have increased sistent effects. By contrast, the many the importance of cash for economic other instances of smaller banking transactions.30 These structural changes crises, which can be seen as small spikes in the composition of GDP should in the respective D-C ratios (e.g. the have led to a secular increase in cur- banking crisis of 1912 in Austria- rency in circulation relative to nomi- Hungary, see Jobst and Rieder in this nal GDP over the course of the 19th volume), did not have persistent effects. century.

29 The fall in the D-C ratio in the U.S.A. after 1960 is due to a shift to M3 deposits and to nonbank financial assets. 30 In the Austrian part of the Habsburg monarchy, the share of agricultural population in the total population was estimated at 75% in 1790 and 48% in 1910. In Hungary the decline in agricultural population started later and was less pronounced (Good, 1984, p. 49).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 109 Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

• During the same period, substantive only a small part of currency in domestic innovations in cashless payments circulation (abstracting from currency occurred. At the beginning of 19th that is circulating abroad) is used for century, bills of exchange (a cash transactions (e.g. Feige, 2012). Esti- substitute) were important instruments mates, partly derived from surveys or in international payments. We do not indirect methods, show that cash held have quantitative evidence on the by households and companies to settle domestic circulation of bills of day-to-day transactions accounts for at exchange, but consider it very likely most one-third of the volume of cur- that their use increased over the 19th rency in circulation. In other words, century. At the same time, the intro- the greater part of cash demand is duction of clearing in large-value driven by developments that might not stock exchange and interbank trans- react to payment innovations. Domestic actions should have increased the hoarding after WW I and WW II, for possibilities for noncash domestic instance, could have been driven by the transactions even further (Kernbauer, experience of financial crisis and political 2016). In 1893, the Austro-Hungarian catastrophes which eroded political and Bank forced banks to open a transac- societal institutions and the public’s tion account to access the central trust in them. Shortly after WW II, a bank’s lending facilities. On the retail certain percentage of deposits were side, the financial innovations in confiscated, which could have further retail banking that started around undermined confidence in banks (Beer, 1890 democratized cashless payments. Gnan and Valderama, 2016). It could These developments should have re have taken decades to restore confi- duced the demand for cash. dence, which would explain the lagging There are many other secular changes D-C ratios in Germany and Austria rel- that could be put forth for which it is ative to the U.S.A. Apart from hoard- rather difficult to find quantitative evi- ing, a significant share of the demand dence. For example, a changing verti- for cash might be related to shadow cal integration of the production chain economic activities (e.g. Feige, 2012). (i.e. more production steps and the use A further reason might be that some of more intermediate inputs) could transactions traditionally done in cash, have either increased or decreased cash e.g. the payments to workers and sup- demand (the direction is not clear a pliers in the construction sector, con- priori), while closer international inte- tinue to be made in this way irrespec- gration should have decreased cash tive of the increasing range of cashless demand. In addition, strong per capita alternatives. income growth after WW II enlarged As a final point, we note that cash the middle-income group, which in has advantages for consumers that turn should have increased cash demand. might not have changed over time. The establishment of a comprehensive Therefore, the proliferation of cashless social security system after WW II also payments could have been partly offset reduced the need to hold precautionary by the public’s continued desire to hold balances, which should have had a and to use cash. Results show that in dampening effect on cash demand, etc. 2011 cash still accounted for the majority Regarding the second major expla- of retail payments in Austria (and nation for the stability of cash demand Germany), both in terms of value and relative to GDP, it should be noted that in terms of the number of payments. In

110 OESTERREICHISCHE NATIONALBANK Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

countries that have moved farther confidence in financial institutions toward the use of cashless payments and increased overall uncertainty. This like the U.S.A. or Canada, cash still also suggests that transactions might accounts for almost 50% of the number not be the most important factor deter- of transactions but only about one- mining demand for cash after all. fourth of their value (Bagnall et al, 2016). It is difficult to find one single 4 Conclusions explanation for these cross-country This paper describes how, and by which differences and we suspect that the high institution, cash was provided over the volume of per capita cash holdings in past 200 years in Austria. Moreover, Austria can be traced to a combination the aim has been to provide a descrip- of factors: (1) the density of bank tive account of the circulation of cash branches and ATMs is high in Austria, from an economic perspective. such that transaction costs for acquiring Results reveal remarkable stability cash are low; (2) Austria is a safe country of demand for currency relative to and people consider it not as dangerous GDP. In response to predictions about as in other countries to keep a larger the imminent demise of cash, one could sum of cash at home; (3) low inflation cite, admittedly somewhat provoca- rates after the pegging of the Austrian tively, that currency in circulation schilling to the Deutsche mark in the amounted to around 8% of GDP in the 1970s decreased the opportunity costs first half of the 19th century, but was of holding cash; (4) cash is most useful still around 7% at the end of the 20th for expenditure control purposes (von century. In recent years, demand for Kalckreuth, Schmidt and Stix, 2014) euro cash has even increased to around and this feature (i.e. the avoidance of 10% of GDP. How can this observation debt) could be more important for con- be reconciled with the vast structural sumers in Austria (and Germany) than changes that have occurred over these in other economies. two centuries and more recently with Evidently, some of these character- the enormous technological advances in istics also apply to countries where people payment technologies? use less cash. However, in combination, First, we have shown that demand they might have established a path for cash is not stable relative to close dependence in the use of cash such that substitutes, e.g. deposits. The massive the importance of cash is only slowly innovations in financial intermediation, decreasing over time, despite massive i.e. the availability of bank deposits and technological advances. Merchants e.g. of cashless giro transactions, have led might have a lower incentive to accept to a sharp rise in deposit-to-currency payment cards if most consumers want ratios. This implies that the importance to transact in cash. A low level of card of cash as a store of wealth has decreased acceptance, in turn, reinforces the role over time. Second, we conjecture that of cash for consumers. the stability of cash relative to GDP is Finally, we note that the recent dis- the outcome of forces that balance each ruptions of the Great Recession have other out. For example, changes in how led even to renewed demand for cash in the economy was organized in the 19th many economies. Evidence in Jobst and century led to increasing demand for Stix (2016) shows that this increase cash, while innovations in financial cannot be assigned to lower interest intermediation have reduced demand. rates alone but also reflects a loss of Moreover, the greater part of demand

MONETARY POLICY & THE ECONOMY Q3– Q4/16 111 Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

for cash might not be susceptible to the 1930s: “After all, the major virtue payment innovations, as it is driven by of cash as an asset is its versatility. It in- hoarding. This role might not have volves a minimum of commitment and changed much over the past 100 years. provides a maximum of flexibility to Two observations can be made as meet emergencies and to take advan- regards the future of cash. First, only tage of opportunities. The more uncer- major financial innovations exert a sub- tain the future, the greater the value of stantial effect on cash demand (relative such flexibility and hence the greater to GDP). This assessment is based on the demand for money is likely to be.” those periods during which a sharp We think this assessment explains why decline in cash demand can be ob- cash will continue to play an important served. These episodes occurred from role in the foreseeable future – at least around 1890 to 1914 and after World in many economies. War II, and in both cases are related to Finally, we have to acknowledge the spread of bank accounts and the that some of our assessments are not possibility of moving funds without the based on hard facts. The lack of knowl- use of cash. Given that cash demand for edge about how the public uses cash is transactions accounts for only a small pertinent to all cash studies – most no- share of overall demand for currency, tably when it comes to the use of cash we expect that the replacement of cash as store of value, which is the key de- by payment cards and by other innova- terminant of total cash demand. Very tive payment methods will lead to only little is known, for instance, about the a rather slow decline in the importance individual motives behind the high per of cash. capita holdings of cash observed in the Second, patterns of demand for euro area today or about the distribu- cash over the past 200 years show that tion of these holdings both within and the development is not linear: major across borders. This lack of knowledge political disruptions, i.e. times of ele- is even more blatant when it comes to vated economic uncertainty, and sys- the more distant past. A proper estima- temic banking crises have had a strong tion of coins in circulation for the 19th and persistent effect on the use of cash. century, including coin migration, is Similarly, the recent increases in cash needed here as much as studies on the demand in many economies might be evolving monetization of the economy. related to the recent financial and eco- To close these gaps in our knowledge, nomic crisis (Jobst and Stix, 2016). more research is needed and, given the This would concur with the observa- enduring importance of cash, is well tion of Friedman and Schwartz (1963, justified. p. 673) regarding the cash demand in

References Argawal R. and M. Kimball. 2015. Breaking through the zero lower bound. IMF Working Paper 15/224. Bagnall, J., D. Bounie, K. P. Huynh, A. Kosse, T. Schmidt, S. Schuh and H. Stix. 2016. Consumer Cash Usage and Management: A Cross-Country Comparison with Diary Survey Data. Forthcoming in: International Journal of Central Banking. Beer, C., E. Gnan and U. Bichler (eds.). 2016. Cash on trial. SUERF conference proceedings 2016/1.

112 OESTERREICHISCHE NATIONALBANK Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

Beer, C., E. Gnan and M. Valderrama. 2016. A (not so brief) history of inflation in Austria. In this volume. Brown, M. and Stix, H. 2015. The Euroization of Bank Deposits in Eastern Europe. In: Economic Policy 81. 1–45. Butschek, F. 1996. Statistische Reihen zur österreichischen Wirtschaftsgeschichte. Die österreichische Wirtschaft seit der Industriellen Revolution. WIFO. Vienna. Cagan, P. 1965. Determinants and effects of changes in the stock of money. NBER. Columbia University Press. Capie, F. and A. Webber. 1985. A Monetary History of the United Kingdom, 1870–1982. Routledge. Compass. 1915. Finanzielles Jahrbuch für Österreich-Ungarn. Vienna. Cvrcek, T. 2013. Wages, Prices, and Living Standards in the Habsburg Empire, 1827–1910. In: Journal of Economic History 73(1). 1–37. Czörnig, C. 1861. Statistisches Handbüchlein für die österreichische Monarchie. Vienna. Der Tresor. 1872–1918. Revue. Statistik und Archiv für Volkswirtschaft und Finanzwesen. Vienna. Deutsche Bundesbank. 1976. Deutsches Geld- und Bankwesen in Zahlen. 1876–1975. Frankfurt: Knapp. Deutsche Bundesbank. 1998. 50 Jahre Deutsche Mark. Monetäre Statistiken von 1948 bis 1997. CD-ROM. Munich: Beck. ECB. 2007. How the Euro Became Our Money. A Short History of the Euro Banknotes and Coins. Frankfurt. ECB. 2015. The international role of the euro. July. Frankfurt. Ertl, H. 1989. Die Wiener Münzstätte in unserer Zeit. Das Hauptmünzamt bzw. die Münze Österreich AG. In: Koch, B. (ed.). Die Wiener Münze. Eine Geschichte der Münzstätte Wien. Österreichische Numismatische Gesellschaft. 113–128. Feige, E. 2012. New Estimates of US Currency Abroad, the Domestic Money Supply and the Unreported Economy. In: Crime, Law and Social Change 57 (3). 239–263. Financial Times. 2015. Scrap cash altogether, says Bank of England’s chief economist. September 18. www.ft.com/cms/s/0/7967908e-5ded-11e5-9846-de406ccb37f2.html (retrieved on July 10, 2016). Friedman, M. and A. J. Schwartz. 1963. A Monetary History of the United States, 1867–1960. Princeton University Press for NBER. Good, D. 1984. The Economic Rise of the Habsburg Empire 1750–1914. Berkeley: University of California Press. Jobst, C. and H. Kernbauer. 2016. The Quest for Stable Money. Central Banking in Austria, 1816–2016. Campus. Frankfurt. New York. Jobst, C. and K. Rieder. 2016. Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931. In this volume. Jobst, C. and T. Scheiber. 2014. Monetary and Economic Statistics for Austria-Hungary: 1863 to 1914. In: Bank of Greece (ed.). South-Eastern European Monetary and Economic Statistics from the Nineteenth Century to World War II. 50–100. Jobst, C. and H. Stix. 2016. The surprising return of cash in an international and historical perspective. Mimeo. Kausel, A. 1979. Österreichs Volkseinkommen 1830–1913. In: Österreichisches Statistisches Zentralamt (ed.). Geschichte und Ergebnisse der zentralen amtlichen Statistik. Vienna. 689–720. Kausel, A., H. Seidel and N. Nemeth. 1965. Österreichs Volkseinkommen 1913–1963. In: Sonderheft der WIFO-Monatsberichte 14. Vienna. Kernbauer, H. 2016. Cashless payments in Austria: the role of the central bank. In this volume.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 113 Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

k. k. Finanzministerium. 1892. Statistische Tabellen zur Währungs-Frage der Österreichisch- ungarischen Monarchie. Vienna. Koch, A. and D. Schneeberger. 2012. Euro cash in Austria, ten years on. In: Monetary Policy & the Economy Q1/12. 28–40. Koch, B. and H. Jungwirth. 1989. Österreichische Münzstätten. In: Koch, B. (ed.). Die Wiener Münze. Eine Geschichte der Münzstätte Wien. Österreichische Numismatische Gesellschaft. 165–172. Komlos, J. 1983. The Diffusion of Financial Technology into the Austro-Hungarian Monarchy towards the End of the Nineteenth Century. In: Komlos, J. (ed.). Economic Development in the Habsburg Monarchy in the Nineteenth Century, Essays. New York: Columbia University Press. 137–163. Komlos, J. 1987. Financial Innovation and the Demand for Money in Austria-Hungary, 1867–1913. In: Journal of Economic History 16. 587–605. Kranister, W. (ed). 1989. Die Geldmacher international. Vom Gulden zum Schilling. Vienna: Staatsdruckerei. Krsnakova, L. and M. Oberleithner. 2012. How Euro Banknotes in Circulation Affect Intra-Eurosystem Balances. In: Monetary Policy & the Economy Q1/12. OeNB. 70–80. Lucam, W. 1861. Die Oesterreichische Nationalbank und ihr Verhältniss zu dem Staate. Vienna: Braumüller. Lucam, W. 1876. Die Oesterreichische Nationalbank während der Dauer des dritten Privilegiums. Vienna: Manz. Menger, C. 1892. Die Valuta-Regulierung in Österreich-Ungarn. In: Jahrbücher für Nationalökonomie und Statistik. Metcalf, C., A. Redish and R. Shearer. 1998. New Estimates of the Canadian Money Stock, 1871–1967. In: The Canadian Journal of Economics / Revue canadienne d'économique 1/31. 104–124. Mühlpeck, V., R. Sandgruber and H. Woitek. 1979. Index der Verbraucherpreise 1800–1914. In: Österreichisches Statistisches Zentralamt (ed.). Geschichte und Ergebnisse der zentralen amtlichen Statistik. Vienna. 649–688. Nationalbank Act. 1922. Federal Law Gazette 823. November 14. Nationalbank Act. 1955. Federal Law Gazette No. 184/1955. OeNB. 1818–1938. 1957–2015. Annual reports of the Nationalbank/Oesterreichisch-ungarische Bank/OeNB. OeNB. 1923–1938. 1946–1989. Mitteilungen des Direktoriums der Oesterreichischen Nationalbank (Communucations of the Governing Board). OeNB. 1990–2004. Statistisches Monatsheft. Vienna. Osili, U. O. and A. Paulson. 2014. Crises and Confidence: Systemic Banking Crises and Depositor Behavior. In: Journal of Financial Economics 111. 646–660. Prammer, D., L. Reiss and W. Köhler-Töglhofer. 2016. The financial relations between the Nationalbank and the government. In this volume. Pressburger, S. 1959. Das österreichische Noteninstitut 1816–1866: Die Vorgeschichte, 1816–1862. Vienna: OeNB. Rieder, R. 2011. Die Entwicklung des Münzsystems im österreichischen Kaisertum und in der Österreichisch-ungarischen Monarchie. Dissertation. Vienna. Rogoff, K. 2014. Costs and Benefits to Phasing Out Paper Currency. NBER Working Paper 20126.

114 OESTERREICHISCHE NATIONALBANK Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

Sandgruber, R. 1982. Die Anfänge der Konsumgesellschaft. Konsumgüterverbrauch, Lebens- standard und Alltagskultur in Österreich im 18. und 19. Jahrhundert. Vienna: Verlage für Geschichte und Politik. Schautzer, A. 2007. Cash logistics in Austria and the Euro Area. In: Monetary Policy & the Economy Q1/07. 138–149. Scheidemünzengesetz (Austrian Coinage Act). 1988. Federal Law Gazette No. 597/1988. Last amended by Federal Law Gazette No. 13/2016. Tafeln zur Statistik der oesterreichischen Monarchie. 1828–1871. Vienna. The Economist. 2007. The end of the cash era. February 15. www.economist.com/node/8702890 (retrieved on July 10, 2016). von Kalckreuth, U., T. Schmidt and H. Stix. 2014. Using Cash to Monitor Liquidity: Implications for Payments, Currency Demand, and Withdrawal Behavior. In: Journal of Money, Credit and Banking 46 (8). December. 1753–1785. Wiener Börse. 1786–1858. Wechsel-Cours in Wien. Vienna. Wiener Börse. 1858–1861. Wechsel-Course an der K. K. Öffentlichen Börse in Wien. Vienna. Wiener Börse. 1861–1875. Cours-Blatt des Gremiums der k. k. Börse-Sensale. Vienna. Willis, H. P. 1896. The Vienna Monetary Treaty of 1857. In: Journal of Political Economy 4(2), 187–207. Wolman, D. 2013. The End of Money: Counterfeiters, Preachers, Techies, Dreamers and the Coming Cashless Society. Da Capo Press. First Trade Paper Edition. Zipser, W. 1997. Auf der Suche nach Stabilität. Das Zentralbankgeldangebot der österreichischen Nationalbank 1923 bis 1937. Frankfurt: Peter Lang.

Data annex and sources of data U.S.A.: 1915–1970: Historical Sta- used in charts tistics of the United States, Colonial Currency in circulation Times to 1970, Washington, D.C., Banknotes: 1820–1860: Lucam (1861); 1975. Series X-414. 1971–2014: Haver 1861–1863: Lucam (1876); 1863–2015: Analytics, Series S111FM1@G10 (Money OeNB (1818–1938, 1957–2015). Stock: M1 SA). Due to unavailability of Government paper money: 1820– M1 before 1915, we have constructed a 1853: Czörnig (1861); 1866–1906: Jobst M1 series for this period by linking and Scheiber (2014). the M1 series with annual growth rates Coins: 1847 and 1867: see running of M2 (source for M2: 1867–1946: text; 1892–1913: Compass (1915); Anderson, Richard G. 2003. Some Tables 1909 is missing; 1925–2001: OeNB of Historical U.S. Currency and Mone- (1923–1938, 1946–1989) and OeNB tary Aggregates Data, Working Paper (1818–1938, 1957–2015). 2003-006A, Federal Reserve Bank of St. Louis). M1, M2, M3 Austria: 1870–1914: Komlos (1983); Nominal GDP 1923–1937: Zipser (1997); 1946– Nominal GDP: 1870–1813: Jobst and 1995: Butschek (1996); 1995–1999: Scheiber (2014). Data for the bench- OeNB (1990–2004). mark years 1830, 1840, 1850 and 1860 Germany: 1876–1975: Deutsche Bun- were projected backward from 1870 by desbank (1976); 1976–1997: Deutsche using nominal GDP growth rates for Bundesbank (1998). the territory of the later First Austrian

MONETARY POLICY & THE ECONOMY Q3– Q4/16 115 Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

Republic as given in Kausel (1979). The Population years between benchmark years were 1820–1862: Tafeln zur Statistik der linearly interpolated. The resulting es- oesterreichischen Monarchie (1828– timate is a very rough indicator and 1863); 1863–1914: Jobst and Scheiber should be used with great care only. (2014). Years with missing values were 1913–1963: Kausel et al. (1965); linearly interpolated. From 1866 on- 1964–2015: Statistics Austria. ward, data for Hungary are only available for census years. Missing years CPI were again linearly interpolated. 1914– 1820–1913: Mühlpeck et al. (1979); 1918: Missing values for the entire 1914–2015: Statistics Austria. The value monarchy were extrapolated using the for 1947 was corrected for a change in rate of change of the population living currency denomination. in the territory of the later First Austrian Republic. 1918–2015: Statistics Austria.

116 OESTERREICHISCHE NATIONALBANK Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

Table A1 Currency in circulation in Austria(-Hungary) from 1820 to 2015

Year Currency unit Total notes in circulation Coins in circulation Total currency in Nominal GDP Population of which issued by of which gold circulation the government

Million currency units Thousands 1820 Florin CM 217.1 165.2 ...... 30,505 1821 Florin CM 166.7 131.9 ...... 30,848 1822 Florin CM 156.2 107.9 ...... 31,219 1823 Florin CM 134.9 83.9 ...... 31,582 1824 Florin CM 135.9 67.9 ...... 31,975 1825 Florin CM 142.0 59.9 ...... 32,378 1826 Florin CM 130.2 47.9 ...... 32,828 1827 Florin CM 127.2 39.9 ...... 33,212 1828 Florin CM 127.6 31.9 ...... 33,551 1829 Florin CM 135.4 27.9 ...... 34,300 1830 Florin CM 131.9 19.9 ...... 1,843 34,504 1831 Florin CM 143.8 19.9 ...... 1,875 34,781 1832 Florin CM 139.8 19.9 ...... 1,907 34,648 1833 Florin CM 144.9 19.9 ...... 1,941 34,736 1834 Florin CM 155.6 19.9 ...... 1,974 35,048 1835 Florin CM 166.0 15.9 ...... 2,009 35,362 1836 Florin CM 169.3 15.9 ...... 2,044 35,663 1837 Florin CM 162.1 15.9 ...... 2,079 35,879 1838 Florin CM 178.8 11.9 ...... 2,115 36,185 1839 Florin CM 178.4 11.9 ...... 2,152 36,556 1840 Florin CM 175.0 7.9 ...... 2,193 36,950 1841 Florin CM 174.5 7.9 ...... 2,267 35,551 1842 Florin CM 178.3 4.9 ...... 2,343 35,804 1843 Florin CM 184.3 4.9 ...... 2,423 35,593 1844 Florin CM 202.6 4.9 ...... 2,504 36,294 1845 Florin CM 218.6 4.9 ...... 2,589 37,037 1846 Florin CM 218.6 4.9 ...... 2,676 37,443 1847 Florin CM 223.9 4.9 120 . . 343.9 2,767 37,780 1848 Florin CM 227.9 4.9 ...... 2,860 37,272 1849 Florin CM 296.5 46.0 ...... 2,957 36,765 1850 Florin CM 343.6 88.2 ...... 3,074 36,258 1851 Florin CM 363.2 147.6 ...... 3,183 35,751 1852 Florin CM 339.0 144.1 ...... 3,296 36,100 1853 Florin CM 312.3 124.6 ...... 3,414 36,435 1854 Florin CM 383.5 x ...... 3,535 36,657 1855 Florin CM 377.9 x ...... 3,661 36,922 1856 Florin CM 380.2 x ...... 3,791 37,186 1857 Florin ö.W. 383.5 x ...... 3,926 37,451 1858 Florin ö.W. 390.5 x ...... 4,065 37,494 1859 Florin ö.W. 466.8 x ...... 4,210 37,537 1860 Florin ö.W. 474.9 x ...... 4,386 37,580 1861 Florin ö.W. 468.9 x ...... 4,475 37,624 1862 Florin ö.W. 426.9 x ...... 4,565 37,667 1863 Florin ö.W. 396.7 x ...... 4,657 37,710 1864 Florin ö.W. 375.8 x ...... 4,751 37,884 1865 Florin ö.W. 351.1 x ...... 4,847 38,137 1866 Florin ö.W. 463.7 179.7 ...... 4,944 35,600 1867 Florin ö.W. 543.9 296.8 8.3 . . 552.1 5,044 35,701 1868 Florin ö.W. 577.5 301.3 ...... 5,146 35,803 1869 Florin ö.W. 598.7 315.0 ...... 5,249 35,904 1870 Florin ö.W. 642.4 345.5 ...... 5,366 36,084 1871 Florin ö.W. 688.1 370.8 ...... 5,733 36,264 1872 Florin ö.W. 689.6 371.2 ...... 6,169 36,444 1873 Florin ö.W. 700.7 341.8 ...... 6,174 36,624 1874 Florin ö.W. 635.7 342.0 ...... 6,264 36,804 1875 Florin ö.W. 628.3 342.1 ...... 6,084 36,984 1876 Florin ö.W. 648.0 352.0 ...... 5,955 37,164 1877 Florin ö.W. 623.2 340.9 ...... 6,316 37,344 1878 Florin ö.W. 649.6 360.8 ...... 6,217 37,523 1879 Florin ö.W. 626.2 309.4 ...... 6,095 37,703 1880 Florin ö.W. 646.5 317.9 ...... 6,308 37,883 1881 Florin ö.W. 670.9 316.7 ...... 6,511 38,231 1882 Florin ö.W. 712.0 343.3 ...... 6,757 38,578 1883 Florin ö.W. 726.5 346.1 ...... 6,820 38,926 1884 Florin ö.W. 719.4 343.7 ...... 6,928 39,273 1885 Florin ö.W. 692.9 329.3 ...... 6,641 39,621 1886 Florin ö.W. 708.5 336.8 ...... 6,400 39,969 1887 Florin ö.W. 722.9 331.8 ...... 6,861 40,316 1888 Florin ö.W. 751.7 326.0 ...... 6,851 40,664 1889 Florin ö.W. 781.5 346.8 ...... 6,855 41,011

MONETARY POLICY & THE ECONOMY Q3– Q4/16 117 Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

Table A1 (continued) Currency in circulation in Austria(-Hungary) from 1820 to 2015

Year Currency unit Total notes in circulation Coins in circulation Total currency in Nominal GDP Population of which issued by of which gold circulation the government

Million currency units Thousands 1890 Florin ö.W. 798.3 352.4 ...... 7,224 41,359 1891 Florin ö.W. 819.5 364.2 ...... 7,470 41,763 1892 Florin ö.W. 807.0 329.0 87.1 3.1 894.0 7,198 42,168 1893 Florin ö.W. 810.1 323.5 102.2 4.1 912.3 7,437 42,573 1894 Florin ö.W. 784.7 276.8 144.3 3.9 929.0 7,589 42,977 1895 Florin ö.W. 785.9 166.0 160.1 4.5 945.9 8,138 43,382 1896 Florin ö.W. 786.6 126.9 160.1 5.6 946.7 7,808 43,787 1897 Florin ö.W. 817.9 118.0 162.6 5.6 980.5 7,769 44,192 1898 Florin ö.W. 850.9 113.4 165.9 7.6 1,016.8 8,291 44,596 1899 Florin ö.W. 833.9 104.9 166.3 8.6 1,000.2 8,726 45,001 1900 Crown 1,579.8 85.8 363.9 18.8 1,943.7 17,595 45,406 1901 Crown 1,608.3 23.3 456.1 117.4 2,064.4 17,217 45,811 1902 Crown 1,639.4 4.2 520.3 204.8 2,159.7 17,986 46,216 1903 Crown 1,773.7 2.9 565.2 239.6 2,338.9 18,674 46,621 1904 Crown 1,754.0 2.7 593.8 265.4 2,347.8 18,451 47,027 1905 Crown 1,849.6 2.6 686.1 350.5 2,535.7 20,987 47,432 1906 Crown 1,984.6 2.5 668.9 320.7 2,653.5 22,659 47,837 1907 Crown 2,028.0 x 668.0 308.9 2,696.0 23,886 48,243 1908 Crown 2,112.9 x 694.0 296.1 2,806.9 24,502 48,648 1909 Crown 2,188.0 x ...... 25,273 49,053 1910 Crown 2,375.9 x 678.5 236.1 3,054.4 26,782 49,458 1911 Crown 2,541.0 x 719.0 253.9 3,260.0 28,299 49,864 1912 Crown 2,815.8 x 743.1 234.8 3,558.9 30,388 50,269 1913 Crown 2,493.6 x 710.1 225.6 3,203.7 30,112 50,674 1914 Crown 5,136.7 x ...... 50,155 1915 Crown 7,162.4 x ...... 49,635 1916 Crown 10,888.6 x ...... 49,115 1917 Crown 18,439.7 x ...... 48,596 1918 Crown 35,588.6 x ...... 48,076 1919 Crown 54,464.6 x ...... 6,420 1920 Crown 30,645.7 x ...... 6,455 1921 Crown 174,114.7 x ...... 6,504 1922 Crown 4,080,177.2 x ...... 6,528 1923 ATS 494.8 x ...... 6,543 1924 ATS 582.5 x ...... 9,257 6,562 1925 ATS 890.0 x 60.4 . . 950.4 10,296 6,582 1926 ATS 947.3 x 76.8 . . 1,024.1 10,283 6,603 1927 ATS 1,005.3 x 76.8 . . 1,082.2 11,110 6,623 1928 ATS 1,067.4 x 89.2 . . 1,156.5 11,678 6,643 1929 ATS 1,094.4 x 92.9 . . 1,187.3 12,087 6,664 1930 ATS 1,090.1 x 91.8 . . 1,181.9 11,560 6,684 1931 ATS 1,183.3 x 97.8 . . 1,281.1 10,360 6,705 1932 ATS 913.8 x 96.8 . . 1,010.5 9,550 6,725 1933 ATS 952.4 x 96.2 . . 1,048.6 9,020 6,746 1934 ATS 963.9 x 108.6 . . 1,072.5 8,980 6,760 1935 ATS 975.6 x 126.2 . . 1,101.8 9,140 6,761 1936 ATS 944.2 x 124.3 . . 1,068.6 9,316 6,758 1937 ATS 943.9 x 125.0 . . 1,068.9 9,822 6,755 1945 ATS 3,740.5 x ...... 6,799 1946 ATS 5,656.5 x ...... 22,847 7,000 1947 ATS 4,325.8 x 74.3 . . 4,400.0 27,764 6,971 1948 ATS 5,634.8 x 111.7 . . 5,746.5 32,111 6,953 1949 ATS 5,721.0 x 123.2 . . 5,844.2 41,606 6,943 1950 ATS 6,348.8 x 129.1 . . 6,477.9 51,993 6,935 1951 ATS 8,032.0 x 151.2 . . 8,183.3 69,190 6,936 1952 ATS 9,048.2 x 203.0 . . 9,251.3 80,128 6,928 1953 ATS 10,474.4 x 280.9 . . 10,755.3 82,652 6,933 1954 ATS 12,252.5 x 287.9 . . 12,540.4 93,574 6,940 1955 ATS 13,026.3 x 314.9 . . 13,341.2 107,296 6,947 1956 ATS 14,259.5 x 338.5 . . 14,597.9 119,303 6,952 1957 ATS 15,402.6 x 432.7 . . 15,835.2 132,068 6,966 1958 ATS 16,598.2 x 699.3 . . 17,297.6 137,178 6,987 1959 ATS 17,692.9 x 692.5 . . 18,385.4 145,900 7,014 1960 ATS 18,726.8 x 699.0 . . 19,425.9 162,893 7,048 1961 ATS 20,878.2 x 778.3 . . 21,656.5 180,726 7,086 1962 ATS 22,419.3 x 797.9 . . 23,217.3 192,134 7,130 1963 ATS 23,970.4 x 847.8 . . 24,818.2 207,083 7,176 1964 ATS 25,740.4 x 906.2 . . 26,646.5 226,730 7,224 1965 ATS 27,547.2 x 966.9 . . 28,514.1 246,491 7,271

118 OESTERREICHISCHE NATIONALBANK Florin, crown, schilling and euro: an overview of 200 years of cash in Austria

Table A1 (continued) Currency in circulation in Austria(-Hungary) from 1820 to 2015

Year Currency unit Total notes in circulation Coins in circulation Total currency in Nominal GDP Population of which issued by of which gold circulation the government

Million currency units Thousands 1966 ATS 29,605.7 x 1,027.1 . . 30,632.7 268,532 7,322 1967 ATS 31,239.6 x 1,091.1 . . 32,330.7 285,593 7,377 1968 ATS 32,449.8 x 1,205.9 . . 33,655.7 306,833 7,415 1969 ATS 34,120.8 x 1,284.3 . . 35,405.1 335,000 7,441 1970 ATS 35,665.5 x 1,287.4 . . 36,952.9 375,885 7,467 1971 ATS 38,998.2 x 1,482.0 . . 40,480.1 419,624 7,501 1972 ATS 44,730.0 x 1,681.7 . . 46,411.7 479,544 7,544 1973 ATS 48,857.0 x 1,889.9 . . 50,746.8 543,458 7,586 1974 ATS 52,365.1 x 2,624.0 . . 54,989.2 618,563 7,599 1975 ATS 56,035.9 x 1,858.2 . . 57,894.2 656,116 7,579 1976 ATS 58,862.4 x 2,034.6 . . 60,897.1 724,745 7,566 1977 ATS 62,194.5 x 2,209.4 . . 64,403.9 843,169 7,568 1978 ATS 67,399.7 x 2,389.3 . . 69,789.0 891,817 7,562 1979 ATS 71,984.7 x 2,624.1 . . 74,608.8 978,249 7,549 1980 ATS 76,795.9 x 2,943.6 . . 79,739.4 1,050,724 7,549 1981 ATS 77,730.7 x 3,194.8 . . 80,925.5 1,119,680 7,569 1982 ATS 80,536.1 x 3,402.4 . . 83,938.5 1,203,367 7,576 1983 ATS 88,676.2 x 3,632.7 . . 92,309.0 1,282,178 7,567 1984 ATS 89,889.6 x 3,833.9 . . 93,723.4 1,346,333 7,571 1985 ATS 90,485.4 x 4,039.8 . . 94,525.2 1,420,861 7,578 1986 ATS 93,900.2 x 4,240.5 . . 98,140.7 1,496,456 7,588 1987 ATS 98,387.0 x 4,466.1 . . 102,853.1 1,553,728 7,598 1988 ATS 103,725.0 x 4,722.9 . . 108,447.9 1,629,442 7,615 1989 ATS 112,761.1 x 5,055.6 . . 117,816.7 1,742,967 7,659 1990 ATS 119,263.8 x 5,452.2 . . 124,716.1 1,873,265 7,729 1991 ATS 127,534.8 x 5,832.4 . . 133,367.2 2,008,306 7,755 1992 ATS 135,004.5 x 6,172.6 . . 141,177.1 2,121,684 7,841 1993 ATS 143,215.9 x 6,537.9 . . 149,753.8 2,191,666 7,906 1994 ATS 151,449.9 x 6,890.6 . . 158,340.5 2,300,980 7,936 1995 ATS 161,412.8 x 7,194.6 . . 168,607.3 2,405,221 7,948 1996 ATS 169,224.4 x 7,503.5 . . 176,727.9 2,506,201 7,959 1997 ATS 171,125.4 x 7,720.5 . . 178,845.9 2,591,198 7,968 1998 ATS 168,822.6 x 7,886.0 . . 176,708.6 2,694,655 7,977 1999 ATS 184,389.0 x 8,092.2 . . 192,481.2 203,418 7,992 2000 ATS 193,098.0 x 8,350.2 . . 201,448.2 213,196 8,012 2001 ATS 141,995.6 x 7,049.3 . . 149,045.0 220,096 8,042 2002 EUR 10,237.0 x ...... 226,303 8,082 2003 EUR 11,691.0 x ...... 230,999 8,118 2004 EUR 13,416.0 x ...... 241,505 8,169 2005 EUR 15,128.0 x ...... 253,009 8,225 2006 EUR 16,815.0 x ...... 266,478 8,268 2007 EUR 18,053.0 x ...... 282,347 8,301 2008 EUR 20,297.0 x ...... 291,930 8,337 2009 EUR 20,640.0 x ...... 286,188 8,341 2010 EUR 21,492.0 x ...... 294,628 8,361 2011 EUR 22,687.0 x ...... 308,630 8,389 2012 EUR 23,298.0 x ...... 317,056 8,426 2013 EUR 24,497.0 x ...... 322,878 8,477 2014 EUR 26,237.0 x ...... 329,296 8,544 2015 EUR 27,795.0 x ...... 337,162 8,585

Source: See data annex. Note: For conversion rates between different currency units see Jobst and Kernbauer (2016). x = No government notes in circulation this year. . . = No data available. Numbers in italics: Authors’ estimates and interpolations.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 119 Cashless payments in Austria: the role of the central bank

Hans Kernbauer1 Austria was a latecomer in the use of cashless payment techniques. This can be explained by the economic backwardness of the Habsburg monarchy compared with Northwestern European countries in the first decades of the 19th century, by the low degree of monetization and the abundance of paper money as a consequence of financing major military campaigns through central bank credits, and by the issuing of paper money by the state. The incorporation of the Austrian Postal Savings Bank in 1882 marked a turning point in the spreading of modern financial technologies in Austria. An amendment to the charter of the Austro- Hungarian Bank in 1887 was a precondition for the substantial growth in the number of giro accounts and in the transaction volumes of the central bank in the following years. In the interwar period, Viennese and Austrian banks had to adapt to the potential of a far smaller economic area. The transaction volumes in the payment system, in particular financial transactions at the stock exchange, were severely hit by the Great Depression. The economic climate between the two World Wars was not conducive to the introduction or spreading of new financial technologies. The economic boom in the Second Austrian Republic, the rapid increase of production, employment and hence income resulted in a wider use of banking services. The Oesterreichische Nationalbank (OeNB) acted primarily as the central clearing house for bank-based payments. The spread of electronically based systems in banking changed the payment process as the use of automated teller machines, point-of-sale terminals and credit cards increased. The introduction of the euro in Austria in 1999 had a fundamental influence on the structure of the Austrian payment system, which had to be aligned with the system used in the euro area. The duties of the OeNB as regulator and supervisor of payment systems in Austria were laid down in an amendment to the Nationalbankgesetz (National- bank Act) in 2001. JEL classification: E42, E50, E58 Keywords: economic and banking history of Austria, development of bank-based payment systems, role of the Austrian central bank

For Austrian households, cash payment transferred over large-value systems, is still the preferred choice for settling on the other hand, add up to 90 to 100 transactions: in 2011 more than 82% times GDP in most developed countries of transactions (65% of the transaction (Kahn et al., 2014, p. 1). The dramatic volume) were paid in cash, hardly less increase of the “payment intensity” i.e. than 15 years ago. The use of point-of- the relation between the transaction sale (POS) terminals increased in this volume and nominal GDP during the period from 2.4% to 25.8% in terms last decades has created enhanced re- of volume, whereas in 2011 only 5% sponsibilities for policymakers, reg- of the transaction volume was settled ulators and supervisors, as failures of through credit cards (Mooslechner et payment systems could have dangerous al., 2012, p. 66). knock-on effects within the financial Retail transactions of households sector and beyond. The collapse of Her- are limited by the amount of private statt Bank (1974) and the difficulties consumption, which represents about encountered in September 2001 in the two-thirds of GDP. Annual payments U.S. large-value system Fedwire have

1 Vienna University of Economics and Business, [email protected]. The author would like to thank William Refereed by: Roberds (Federal Reserve Bank of Atlanta) and Clemens Jobst, Helmut Stix, Markus Pammer and Walpurga Köhler-Töglhofer (OeNB) for their valuable suggestions. The author would also like to thank Douglas Deitemyer William Roberds, and Susanne Steinacher for valuable language support. Federal Reserve The views expressed in this paper are exclusively those of the author and do not necessarily reflect those of the Bank of Atlanta OeNB or the Eurosystem.

120 OESTERREICHISCHE NATIONALBANK Cashless payments in Austria: the role of the central bank

demonstrated the liquidity risks and tems in banking from the mid-1970s the risks presented by the nonfinality of onwards revolutionized the payment payments. Most central banks, and in system, too. Wages, salaries and pen- particular the central banks of the euro sions were increasingly paid out cash- area, nowadays bear responsibility for less; cash dispensers, automated teller the smooth operation of payment sys- machines (ATMs) and POS termi- tems. This is seen as a prerequisite for nals – accessed via bankers’ cards and conducting monetary policy and main- credit cards – were used with grow- taining the stability of the financial ing frequency. The OeNB acted as a sector, while prior to the tremendous central clearer in the payment system growth of electronically transferred and was involved in the supervision of funds the tasks of central banks for the the participating institutions (section payment system were more or less re- 4). In preparation for the introduction stricted to the production of banknotes of the euro, the large-value payment and coins and to stocking banks with systems of the participating countries the necessary cash reserves. were modernized and harmonized, as This paper deals with the his- a secure payment system is necessary tory of cashless (or cash-saving) pay- for financial stability in general, and in ments in Austria and the specific role particular for a smooth distribution of of the Oesterreichische Nationalbank base money within the euro area. In ad- (OeNB) in this process. Section 1 deals dition, the mushrooming of the finance with the beginning of giro and clearing sector and – as mentioned above – the transactions in Austria. The low degree concomitant increase of payment inten- of monetization in general and the am- sity created new challenges for operat- ple supply of banknotes in the first de- ing a safe payment infrastructure. Core cades of the 19th century were respon- principles for systemically important sible for the slow start of this business payment systems were developed by a in Austria compared with other West- committee of the Bank for International ern European countries. Encouraged Settlements (BIS). The responsibilities by the Ministry of Finance, the initia- of the OeNB as operator, facilitator and tives of the Nationalbank to propagate supervisor of the Austrian payment sys- giro transactions were successful only tem were defined in more detail in a after the foundation of the Postal Sav- 2001 amendment to the Nationalbank ings Bank in 1882 and changes in the Act (section 5). 1887 central banking law regulating the coverage of banknotes by precious 1 Tentative beginnings of cashless payments in the first half of the metal (section 2). The breakup of the th Habsburg monarchy after World War I 19 century (WW I) and the restructuring and cri- Its rules of procedure dating from 1817 ses of the Austrian banking sector in authorized the Nationalbank, inter alia, the interwar period were not conducive to operate the giro business, although to the spreading and improvement of the relevant charter was very restric- financial technologies (section 3). The tive regarding this line of business: only exuberant economic growth after the shareholders resident in Vienna were stabilization of the Austrian currency at permitted to open a giro account at the the beginning of the 1950s, the strong Nationalbank. Account holders could increase in demand and savings depos- deposit banknotes, silver coins and bills its and the soaring use of electronic sys- of exchange denominated in the cur-

MONETARY POLICY & THE ECONOMY Q3– Q4/16 121 Cashless payments in Austria: the role of the central bank

rency of the bank. The deposit could be and banknotes because money orders used for cash withdrawals or trans- could be executed quicker and safer. ferred to other accounts at the bank The government paid lower fees than (Bubenik, 1888, p. 2). The giro busi- private firms. In the 1830s the National ness was not activated during the first bank’s fee income derived from money privilege of the Nationalbank, which orders amounted to fl. 30,000 to fl. was granted for the period from 1816 40,000 per year (around 1% of net to 1841. Baron Carl Joseph Alois von profits) (Lederer, 1847, p. 174ff). Lederer, Governor of the National- The slow start of cashless payments bank, justified this in retrospect by the and banking in general in Austria was, business community’s lack of interest in as mentioned above, attributable to the bank-based payments (Lederer, 1847, relative economic backwardness of the p. 234). However, the restriction of the Habsburg monarchy compared with giro business to Viennese shareholders the countries in Northwestern Eu- seems to be a more convincing argu- rope. A reliable quantitative compar- ment for the subdued interest of enter- ison of economic levels in Continental prises in banking with the National- Europe in the first decades of the 19th bank. century is not possible for lack of data Right from the beginning of its op- and would in any case have to take into erations, the Nationalbank was active account the huge economic discrepan- in the execution of money orders, i.e. cies among the various provinces of the transfer of cash from the head of- the monarchy. “The vast majority of fice in Vienna to branches2 in provin- producers,” writes Gross (1983) in his cial towns and vice versa. In the period essay on Austria-Hungary’s role in the from 1818 to 1829 the transfer of cash world economy, “were still chiefly en- for private firms from branch offices to gaged in subsistence farming, and the Vienna was about three times as high estimation of this type of output and as the reverse, whereas from 1830 to of peasant productivity is a frustrating 1841 the amounts in both directions exercise even in our age of worldwide were more or less equal. The govern- national income accounting.” (Gross, ment used Nationalbank money orders 1983, p. 3). Only farmers in the vicinity for the shipment of tax receipts from of larger towns were solely market the provinces to the capital: from 1830 producers, and even artisans quite of- to 1841 these transfers averaged florin ten exchanged their products in kind.3 (fl.) 17.8 million per year compared The lack of efficient transport facilities with yearly private money transfers before the building of railways, which to Vienna of fl. 7.4 million (Lederer, began in the 1830s, and the lack of 1847, p. 177). The fees for money or- navigable rivers in Austria defined a ders were calculated according to the rather small zone around larger towns amount transferred and the spatial dis- for marketable production of perishable tance spanned. They were higher than foodstuffs and trading activities in the fees for postal deliveries of coins general.

2 These branches (Filial-Banken) were entitled only to exchange paper money into coins and to execute money orders. In the Habsburg monarchy, there were only 12 such branch offices in operation during the first privilege. They were – except for the one in Prague – part of the financial administration of the state. See Pressburger (1959, p. 168). 3 The greater part of the salaries of farm laborers was paid in kind, too.

122 OESTERREICHISCHE NATIONALBANK Cashless payments in Austria: the role of the central bank

The monarchy’s relative backward- in the development of banking in gen- ness resulted in a low degree of mon- eral and the payment system in par- etization of economic activities, which ticular. The creation of the Bank of correlated with the state of the bank- Amsterdam (Amsterdamsche Wisselbank) ing system.4 The Nationalbank, the in 1609, in some ways emulating Ital- only joint stock bank in Austria until ian banks of the 15th and 16th centuries the mid-19th century, acted primar- (Houtman-De Smedt and van der Wee, ily as the banker of the state and refi- 1993, p. 133), addressed the problem nanced – to a lesser degree – bills of of “incremental debasement” (a term exchange of Viennese private banks. first coined by Adam Smith) faced by Initially, private banking was mainly small states using a metallic monetary an add-on to wholesale trading. Pri- system: growing trading activities ne- vate bankers were primarily engaged cessitated an increase in the supply of in buying public bonds and financing money that was often accomplished by consumption expenditures of the aris- using foreign coins. At the end of the tocracy, although some of them were 16th century, 800 different foreign coins also active in financing industrial ven- were officially recognized in Amster- tures (Matis and Weber, 1993, p. 318ff; dam (Quinn and Roberds, 2006, p. 1), Matis, 2005; Teichova, 1997, pp. 214– not few of them debased, creating “ver- 228). The financial system of the first itable monetary chaos” (Houtman-De half of the 19th century can be summed Smedt and van der Wee, 1993, p. up in the phrase “few banks and many 126ff). An ordinance of the city of Am- banknotes,” as the abundance of paper sterdam declared the settlement of bills money originating from the financing of exchange stemming from commer- of the Napoleonic wars was reduced cial activities obligatory via the Bank of only belatedly (Jobst and Kernbauer, Amsterdam, and merchants were com- 2016, p. 48ff.). It stands to reason that pelled to open an account at this bank. such conditions were not as conducive By these measures the losses of mer- to cash-saving financial innovations as chants stemming from the debasement the use of giro accounts and the clear- of coins were effectively reduced. ing of transactions would have been: In general, the use of bank accounts the business community did not show for the settlement of financial balances any interest in using a bank-based pay- of merchants leads to the establishment ment system. It was only in the last of clearing institutions, which square decade of the 19th century that giro claims and liabilities of banks. The transactions of Austrian banks and the London Bankers’ Clearing House was central bank became (macro)economi- founded as early as 1775. Until 1854, cally important. when joint stock banks were allowed to Internationally, the settlement of participate in the clearing of financial receivables and liabilities stemming balances, the Bankers’ Clearing House from trading activities has been done was active for private banks only. In in a cashless manner since antiquity 1864 the Bank of England joined this (Koch, 1910, p. 15; Körner, 1993, venture, followed by English country p. 66ff). In the Middle Ages interna- banks in 1865 (Rauchberg, 1897, tional trade fairs played a decisive role p. 21). Cashless payments had gained

4 In this context, Komlos (1983, p. 137) speaks of the “symbiotic relationship between the financial infrastructure and the industrial development of an economy.”

MONETARY POLICY & THE ECONOMY Q3– Q4/16 123 Cashless payments in Austria: the role of the central bank

wider currency in England when in of fl. 0.25 per fl. 1 million in 1856, the 1833 joint stock banks started paying turnover dwindled to fl. 30 million to 50 interest on deposits. The establishment million per year (Bubenik, 1888, p. 4ff). of the British Postal Savings Bank in The Nationalbank’s financing of the 1861 boosted the use of giro accounts repression of the 1848 revolution and in England. In the second half of the of the war against Piedmont-Sardinia 19th century some wages and salaries in in 1859 increased the circulation of England were already paid cashless banknotes accordingly (Jobst and (Born, 1993, p. 195). Kernbauer, 2016, p. 73), redressing the The New York Clearing House was inducement to use cash-saving payment founded in 1853; in the absence of a methods (Hammerschlag, 1905, p. 19). central bank, it was to exert a stabiliz- When in 1862 the charter of the bank ing influence on the U.S. banking sec- was renewed along the lines of England’s tor in the years to come. Together with Peel’s Act, restricting the amount of New York banks it financed the money banknotes not covered by silver reserves market and was “pulled into the role of to fl. 200 million, the central bank lender of last resort without benefit of considered it necessary to foster the the privileges of that Bank,” i.e. the clearing of balances between the joint Bank of England (Wood, 2011, p. 134ff; stock banks that had been founded in Goodhart, 1988, p. 29ff). the decade since 1853 (Matis and We- In the Habsburg monarchy, by con- ber, 1993, p. 322ff; März and Socher, trast, the monetization of economic 1973, p. 329ff). In 1864 the Clearing activities and the use of giro accounts Hall (Saldosaal) was instituted at the started, as already mentioned, with Nationalbank to clear the mutual some delay. balances of the Nationalbank, Creditan- stalt, the Niederösterreichische Escompte- 2 Financial innovations in the last th Gesellschaft and the Anglo-Österreichische decades of the 19 century Bank (Bubenik, 1888, p. 8). Although The efforts of the government and the the Nationalbank did not charge any Nationalbank to disseminate bank- fees for its giro and clearing business, based payments in the period since the this initiative was not successful; this beginning of the 1840s were not suc- was not surprising because the financ- cessful. After the renewal of its privi- ing of the war of 1866 against Prussia lege in 1841 the Nationalbank, encour- and Italy swelled the circulation of aged by the government, activated the paper money – banknotes and govern- giro business in 1842. The turnover ment paper money or state notes – sub- of giro accounts increased quickly and stantially (Jobst and Kernbauer, 2016, amounted to about fl. 200 million in p. 81; Rauchberg, 1886, p. 4). the following years. However, the num- In 1871, during the boom period ber of giro accounts never exceeded 26, of the Gründerzeit, the Wiener Giro- und of which 14 were responsible for the Cassen-Verein was founded, primarily for largest part of the transactions. In ad- the clearing of securities transactions. dition, the Nationalbank’s giro business It offered giro accounts and acted as a consisted mainly of the encashment of hub for the encashment of bills of ex- bills of exchange and not of the transfer change and money orders. Its giro busi- of balances between accounts. When ness numbered about 500 participants. the Nationalbank started to charge a fee The turnover increased rapidly from

124 OESTERREICHISCHE NATIONALBANK Cashless payments in Austria: the role of the central bank

fl. 2.3 billion in 1872 to fl. 9.2 billion the macroeconomic requirements for in 1895 but fluctuated widely from year this type of banking business.” 5 to year (Rauchberg, 1897, p. 120ff). In 1888, the Budapest Clearing In 1872, the Clearing Hall was trans- Union was established, encouraged formed into the Wiener Saldierungs- by the Austro-Hungarian Bank, and Verein (Vienna Clearing Union). Its in 1895 clearing houses in Brno and turnover did not show any increase Prague followed. There, transactions in the coming decade and remained remained quite unimportant as long low compared with clearing houses in as the payment process in general was England, Germany, France, Italy and executed in the traditional manner. the United States (Rauchberg, 1897, The clearing houses demonstrated, in p. 18ff; Pressburger, 1962, p. 1120ff). a way, banks’ readiness to encourage The number of banks participating in giro transactions, but gained no practi- the Vienna Clearing Union never ex- cal significance until the 1890s (Rauch- ceeded 14, of which the Nationalbank berg, 1897, p. 17ff). and the Wiener Giro- und Cassen-Verein As already mentioned, the giro (Vienna Giro and Central Securities business of the Austro-Hungarian Bank Depositary Union) contributed half did not really take off until some arti- of the turnover. The Vienna Clearing cles of its charter were changed by a Union was only partially successful in law of May 1887: the compensation of claims and liabil- • At least 40% of notes in circulation ities of the participating banks, as less had to be covered by the metal reserve than a quarter of the turnover could be of the central bank, the remaining balanced, the remaining part being note circulation and demand deposits settled via the Nationalbank (Rauchberg, of the central bank by bills of exchange 1897, p. 5). and lombard loans. The giro business of the Oester • For any amount of banknotes in cir- reichisch-ungarische Bank (Austro-Hun- culation that was not covered by the garian Bank), i.e. the reorganized central metal reserve and exceeded fl. 200 bank as of 1878 (Jobst and Kernbauer, million, the central bank had to pay a 2016, p. 112ff), got a boost at the end 5% banknote tax, but was no longer of the 1880s when legal provisions running the risk of being sanctioned concerning the relation between by having its privilege withdrawn. banknotes in circulation and the • Paper money issued by the state amount of available precious metals (government paper money) and owned were changed (Mecenseffy, 1896, by the central bank reduced the p.46ff). “Until then,” wrote Rauchberg amount of banknotes to be covered (1897, p. 56) in his famous book about by bills of exchange or securities the clearing and giro business of the (Bubenik, 1888, p. 15ff; Zuckerkandl, Austro-Hungarian Bank, “the giro 1899, p. 249). business had a virtual existence only. These legal changes made the enlarge- Besides the small number of partici- ment of the giro business possible with- pants and the insignificant amount of out risking to loose the privilege in the turnover, the structure of the giro case the money supply overruns the organization in the bank did not fulfill legal limit. In addition, it was the new

5 Author’s translation.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 125 Cashless payments in Austria: the role of the central bank

Chart 1 Giro transactions of the Austro-Hungarian Bank ﬂ. million Number 50,000 7,000

45,000 6,000 40,000

35,000 5,000

30,000 4,000 25,000 3,000 20,000

15,000 2,000

10,000 1,000 5,000

0 0 1887 1890 1895 1900 1905 1910 1913 Transactions (left-hand scale) Accounts (right-hand scale)

Source: Mecenseﬀy (1896, p. 48); Oesterreichisch-ungarische Bank (1887–1913).

organization of the central bank’s giro success of the Austrian Postal Savings business which was responsible for the Bank (PSB), founded in 1882, which tremendous growth of giro transac- fostered and popularized the giro busi- tions in the ensuing years. Previously, ness in the Habsburg monarchy, even giro transactions had been possible only for smaller firms and private individuals. locally: exclusively in Vienna and The changes of the charter of the Budapest. From 1888 onward, this Austro-Hungarian Bank and the re- type of business was offered in all organization of its giro business were branches, and transfers between all modeled on the legal provisions of branches were possible without a fee. the German Reichsbank. Right from In 1887, only 16 giro accounts were the start of its activities in 1876, the held at the Austro-Hungarian Bank, Reichsbank had facilitated giro trans- with turnover amounting to fl. 817 actions in all branches, and as of 1883 million. During the following years the in all sub branches as well, thus cre- number of accounts and the volume of ating an empire-wide organization for transactions increased strongly (chart 1). cashless payments. Transactions were The growth of the number of giro free of charge. Beginning in 1883, in- accounts can partly be explained by the stitutions that wanted to sell bills of ruling of the Austro-Hungarian Bank exchange to the Reichsbank or to take in 1893 that any firm intending to sell out lombard loans had to have a giro bills of exchange to the bank had to account at the Reichsbank. It is no have a giro account. Nevertheless, in wonder that giro turnover increased 1896 only 5.8% of registered compa- strongly in the following years, as not nies in the monarchy – 8.2% in Vienna, only private persons and firms but also 9.2% in Budapest – held a giro account the public administration, the postal at the central bank (Rauchberg, 1897, service, the Prussian railways, the mil- p. 57f). This can be explained by the itary and other institutions made use

126 OESTERREICHISCHE NATIONALBANK Cashless payments in Austria: the role of the central bank

of the modern payment system. The exceeded the number of accounts with number of giro accounts of the Reichs the Austro-Hungarian Bank, totaling bank grew from 3,200 in 1877 to more fewer than 6,000. The turnover at the than 12,000 in 1896, while turnover PSB of about fl. 200 million – 400 mil- increased from 27 billion German lion crowns – was much lower than the reichsmark to 106 billion reichsmark giro transactions of the Austro-Hun- over the same period (Rauchberg, 1897, garian Bank, which acted as the cen- p. 83ff; Koch, 1910, p. 14ff). Born tral clearing house for the monarchy: (1993, p. 194) argued that the German in 1913 the volume of central bank giro central bank’s endeavor to stimulate transfers amounted to fl. 47 billion (94 cashless payments could be explained billion crowns) as illustrated in table 1. by the Reichsbank’s “erroneous” be- The substantial increase in the lief that giro money and bank transfers number of PSB giro accounts after 1895 would not increase the money supply was influenced by the government de- and the velocity of money. According cision to allow taxes to be paid through to Mises (1924, p. 397ff), “the main- the PSB network beginning in 1896 in stream of German economic policy Lower Austria and from 1898 onward makers” was convinced that the use of in all provinces (Komlos, 1983, p. 142f; bank money would lead to a strength- Wagner and Tomanek, 1983, p. 108). ening of the central bank’s metal re- The transfer of money between the serves and to a consequential reduction Austro-Hungarian Bank and the PSB of interest rates. But this, maintained network and vice versa was possible as Mises (1924), would be the case only to of 1889 in Austria and as of 1891 in the extent that the added metal reserve Hungary (Rauchberg, 1897, p. 77). was used to raise the supply of capital The government fostered the cash- goods by imports. saving use of giro accounts for mone- In the Habsburg monarchy, the tary policy reasons as well: given the founding of the PSB was probably the fall in the price of silver, the minting of most important trigger for the promul- silver for private individuals was sus- gation of the giro and checking business, because this move introduced Table 1 the new financial technology to people who would otherwise not have opened Giro business of the Austro-Hungarian Bank and the a bank account because their savings Austrian Postal Savings Bank were too small or transactions too in- AHB PSB AHB PSB frequent (k. k. Finanzministerium, 1895, Number of accounts Transactions (fl. million) p. 440). The PSB’s network in the Aus- 1887 19 12,981 817 28 trian part of the monarchy comprised 1890 984 17,808 4,477 34 more than 4,000 branches (Wagner and 1895 2,587 28,363 7,930 55 Tomanek, 1983, p. 39), whereas the 1900 5,101 42,658 12,259 55 1905 5,440 67,804 24,295 154 Austro-Hungarian Bank had only 101 1910 5,610 102,574 41,153 197 6 branches and the commercial banks 1913 5,761 122,870 47,105 196 even fewer by far. It is therefore not Source: Mecenseffy (1896, p. 48); Oesterreichisch-ungarische Bank (1887, 1890, 1895, 1900, 1905, 1910, 1913); Wagner and Tomanek (1983, passim). surprising that in 1910 the more than Note: AHB = Austro-Hungarian Bank; PSB = Postal Savings Bank. 100,000 giro accounts with the PSB far

6 See Jobst and Kernbauer (2016, p. 133). In addition to its own branch network, the Austro-Hungarian Bank used contract offices at financial institutions to discount bills or for the encashment of bills.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 127 Cashless payments in Austria: the role of the central bank

pended in 1878 and the introduction of part in the payment process.” (Rauch- a gold-based currency was considered berg, 1897, p. 147).7 in the following years. To reduce the After the demise of the Habsburg necessary amount of gold to back monarchy, Austria’s financial infra- money in circulation, the substitution structure, which had been modern- of cash by other means of payments was ized in the last decades of the 19th cen- considered advisable. tury, had to be adapted to the economic The immediate success of the postal potential of the much smaller First saving system in collecting the savings Austrian Republic. This happened not of lower-income earners as well and in in a planned way but was forced on the popularizing the use of giro accounts country at a very high price through had positive macroeconomic effects, costly financial crises: the shrinkage too: the substitution of deposits for of an individual economic sector is not cash increased the money multiplier conducive to the introduction of ad- and hence the supply of loanable funds, vanced technologies. and rural areas became accustomed to modern banking services and were 3 Restructuring of the Austrian consequently drawn into the money financial sector in the interwar economy. As expenditures could be period increased beyond the current income, In its 1929 Annual Report, the OeNB effective demand was rising with ac- stated that the failure of Vienna’s sec- celerating effects on the growth rate of ond largest bank, the Bodencreditanstalt, the real sector of the economy (Komlos, was due to the bank’s neglecting to 1983, p. 147f). adapt its business model in a timely In Hungary, the Hungarian Postal manner to the dramatic downsizing of Savings Bank was founded in 1885, Austria. This characterization of the with giro transactions starting four business policy of the Bodencreditanstalt years later. The growth rate of giro can to a certain degree also be applied transactions of this institution was to other Austrian banks. lower than that of the PSB in Austria, The Austrian banking sector was because in Hungary, the Austro-Hun- far too large for the new republic. In garian Bank had already offered giro 1913, banks located in the area of the transfers before the Hungarian Postal later Republic of Austria had accounted Savings Bank was established and was for more than two-thirds of own therefore able to gain a bigger market funds of all banks in Cisleithania, while share in this part of the monarchy the economic potential of the First (Rauchberg, 1897, p. 172). Austrian Republic was hardly more Austrian joint stock banks played than one-quarter of that of the formerly a minor role in giro transactions, as Austrian part of the monarchy (Jobst the few available sources for the last and Kernbauer, 2016, p. 171f). The decades of the 19th century show. Al- First Austrian Republic had to cope though there was a certain increase in with the inherited process of inflation, giro transfers, in the mid-1890s joint which ushered in a period of hyperin- stock banks “were still far away from flation in 1921, and with the alignment using giro transactions for a substantial of the industrial and financial structure

7 Author’s translation.

128 OESTERREICHISCHE NATIONALBANK Cashless payments in Austria: the role of the central bank

to the much reduced economic poten- zentrale der österreichischen Genos- tial of a small state. senschaften (Giro Center of the Austrian When the financial institutions of Cooperatives ) with the aim of equalizing the First Austrian Republic re-evalu- the money balances of agricultural and ated their assets and liabilities after the commercial unions of cooperatives.9 stabilization of the currency and the In the interwar period these institu- introduction of the Austrian schilling tions’ giro transactions were negligible in 1925, they found their own funds (Ebner, 1962, p. 9). reduced by three-quarters compared The function of the OeNB as the with 1913 (Kernbauer, 1991, p. 172ff). central clearing institution of the Aus- The infrastructure of the banking sec- trian payment system was vaguely de- tor was more or less the same as before fined in Article 1 of its charter of 1922. the war, though the business activities Article 1 states that besides other duties were much reduced. the OeNB had to “facilitate the settle With regard to payment systems, ment of payments.”10 This general clause the OeNB, which was re-established in is not detailed further in the charter. In 1922, remained the country’s central particular, a controlling or supervising clearing institution. The central bank’s role of the central bank for the payment payments division handled the balances system is not specified. This is in ac- to be cleared of the PSB, the Wiener cordance with the prevalent paradigm Giro- und Cassen-Verein, the Wiener Sal of this period which held that banks dierungsverein and the larger Austrian as private institutions are not liable to banks. Savings banks commenced their transfer any information about their giro business only in 1937, with the business activities or the structure of founding of the Girovereinigung der their balance sheets to the central bank Sparkassen (Giro Union of the Savings (Jobst and Kernbauer, 2016, p. 170). Banks).8 Efforts to activate the giro Table 2 shows the amount of giro business of savings banks were doomed transactions in Austria in the interwar to failure for various reasons: The lack period. The OeNB as central clearer of a clearing institution for the savings handled the bulk of transactions. Their bank sector was equivalent to the non- development mirrors in a dispropor- existence of a giro network, as savings tionate way the business cycle of the banks could not have branches, and the interwar period: During the Great supervising authorities were restric- Depression the amount of transactions tive in granting a license for this line related to economic activities of the of business. Among other things, the real and the financial sectors shrank amount of giro deposits was restricted dramatically. In 1935 the volume of to 10% of all deposits (Hauptverband transactions was 50% lower than in der österreichischen Sparkassen, 1972, 1929. This decrease was caused pri- p. 540ff). In 1927 the cooperative marily by the slump of the trading vol- banking sector incorporated the Giro ume of securities on the Vienna stock

8 See Hauptverband der österreichischen Sparkassen (1972, p. 821ff.). After Austria’s annexation in March 1938, the Girovereinigung was restructured along the lines of the German Girozentrale; see Giefing and Auracher (1977, p. 10ff). 9 See Girozentrale der österreichischen Genossenschaften (1929) and Klauhs (1977, p. 7–11). The sluggish economic development of the following years was restraining the growth of this institution in the interwar period. 10 Article 1 of the charter reads in German: “Die Oesterreichische Nationalbank […] hat im Gebiete der Republik […] die Zahlungsausgleichungen zu erleichtern.”

MONETARY POLICY & THE ECONOMY Q3– Q4/16 129 Cashless payments in Austria: the role of the central bank

Table 2 1937 from 150% to 270%. The num- Giro transactions 1913 to 1937 ber of PSB giro account holders was OeNB/Austro- Austrian Postal Wiener Wiener more or less stagnant in the interwar Hungarian Bank Savings Bank Giro- und Saldierungs- period (Wagner and Tomanek, 1983, Cassen-Verein verein pp. 224 and 253). The rise in payment ATS billion intensity observed in these years will 1923 14.4 8.9 19.7 1.4 require further research. 1924 21.1 13.5 16.8 2.0 1925 30.0 15.7 14.7 1.9 The Wiener Saldierungsverein had the 1926 41.1 17.4 27.2 2.4 smallest turnover figures among the 1927 43.4 19.2 33.2 3.1 Austrian giro and clearing institutions. 1928 45.0 22.3 31.5 3.5 1929 49.8 24.8 33.5 4.1 Its volume of transactions fell by 50% 1930 49.1 26.1 31.6 4.1 between 1929 and 1934. In 1936, ten 1931 49.7 25.0 23.5 3.1 1932 41.2 22.5 16.1 2.6 financial institutions were members of 1933 36.8 21.4 16.0 2.3 the Wiener Saldierungsverein, among them 1934 26.7 22.7 9.3 2.0 the OeNB, the Wiener Giro- und Cassen- 1935 24.7 24.1 6.4 2.3 1936 26.9 24.6 6.4 2.4 Verein, the PSB and other Viennese banks 1937 28.2 26.6 7.2 2.6 (Szedenik, 1937, p. 134f). Describing the development of Source: Governing Board of the OeNB (various years, passim); Wagner and Tomanek (1983, passim); Wärmer (1936, p. 77). cashless payments in Austria in the interwar period, Ebner (1962, p. 5) reached a downbeat conclusion: “It exchange, which is reflected in the was regrettable and without doubt a big turnover data of the Wiener Giro- und drawback for the Austrian economy Cassen-Verein: when the settlement of that until 1938, basically cashless pay- transactions of securities could not be ments were restricted to the Postal executed by the participating parties in Savings Bank.”11 Compared with the Wiener Giro- und Cassen-Verein, they Germany, the Austrian payment sys- were cleared using the accounts at the tem was underdeveloped. Ebner (1962) OeNB. As can be seen from table 2, the blamed this backwardness on the credit transaction volume of the Wiener Giro- institutions, which, in his opinion, only und Cassen- Verein fell by ATS 24 billion too late recognized the importance of or 81% between 1929 and 1935. cashless payments for their own inter- The giro business of the PSB is ests and for the economy as a whole. directly related to transactions connected to production and consumption 4 The spread of cashless pay- activities in the economy. The turnover ments in a booming economy figures of the PSB’s giro business are The annexation of Austria into the therefore more in line with the devel- German Reich in 1938 forced Austrian opment of the nominal gross national credit institutions to adapt their organi- product (GNP). In 1933 nominal GNP zational structures and their modes of was 25% lower than in 1929, while the doing business to the procedures pre- PSB’s giro business was down by 18% vailing in Germany. Savings banks and (Butschek, 1999, table 5.1). It is inter- cooperative banking institutions were esting to note that the PSB giro trans- integrated into the structures that ex- action volumes as a percentage of nom- isted in the German Reich before 1938: inal GNP went up between 1924 and in particular, the giro payment systems

11 Author’s tranlsation.

130 OESTERREICHISCHE NATIONALBANK Cashless payments in Austria: the role of the central bank

in place in Germany were adopted and the population. The use of giro trans after World War II (WW II) became a actions increased strongly: In 1977 the kind of template for the organization of savings bank sector boasted more than payment systems in Austria. After WW 1.5 million accounts dedicated to giro II, different payment systems came into payments (Sadleder, 1977, p. 25). being throughout the different sections The giro business in the cooperative of the banking industry except for joint banking sector started in earnest only stock banks, which balanced giro trans- after WW II. It benefited from the actions in-house as far as possible and technical and organizational experience cleared the remaining positions by using of the respective German institutions their giro account at the OeNB. of which it had been part during the an- For the savings bank sector, the nexation period (Ebner, 1962, p. 9). Girozentrale der österreichischen Sparkassen Giro transactions in the agricultural AG (Giro Center of Austrian Savings cooperative sector were cleared at the Banks) acted as a clearing house. Par- regional level (Raiffeisen regional co- ticipating savings banks had to conduct operative banks) and a second time at their transactions according to defined the Genossenschaftliche Zentralbank (Co- principles for the giro and checking operative Central Bank). The remain- business and the encashment of bills of ing balances were settled via the latter’s exchange. The flourishing of the Aus- giro account at the OeNB. trian economy after the stabilization There are no data available on trans- of the currency in 1951/52 led to a action volumes for the joint stock and commensurate expansion of real in- cooperative banking sectors.12 The come and savings for a wide stratum of savings bank sector reported a transac-

Chart 2 Giro business of the Austrian Postal Savings Bank Number ATS billion 350,000 5,000

4,500 300,000 4,000

250,000 3,500

3,000 200,000 2,500 150,000 2,000

100,000 1,500

1,000 50,000 500

0 0 1948 1950 1955 1960 1965 1970 1975 1980 Turnover (right-hand scale) Accounts (left-hand scale)

Source: Wagner and Tomanek (1983, passim).

12 The giro business of the provincial mortgage banks was of minor importance.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 131 Cashless payments in Austria: the role of the central bank

tion volume of less than ATS 300 mil- of salaries or pensions amounted to lion in 1969 (Peter, n.d., p. 21), when 61% of the population. Technically, the the PSB had a turnover of ATS 1,322.2 precondition for this development was million (Wagner and Tomanek, 1983, the diffusion of electronic systems in p. 314). One can therefore be quite banking. sure that the PSB was by far the largest In the Second Austrian Republic, as giro payment institution in the Second it had in the interwar period, the OeNB Austrian Republic as well. The transac- acted as the central clearing house for tion volume of the PSB amounted to giro transactions. The legal framework ATS 92.6 billion in 1948 and increased for these activities did not change from to ATS 473.4 billion in 1980 (chart 2). the situation before WW II until an This rise was to some extent the ef- amendment was made to the National fect of the PSB’s function as the banker bank Act in 2001. The turnover of giro of the public administration, including accounts amounted to ATS 437 billion the social security system: Taxes col- in 1956 and to ATS 14,711.1 billion in lected were transferred via PSB giro ac- 1988, when the OeNB stopped pub- counts to the main account of the state lishing such figures (chart 3).13 Turn- managed by the Ministry of Finance, over increased in this period from four and funds were made available by the times to about 10 times the nominal Ministry of Finance for subordinated GNP. entities and other ministries, as were In 1989, Austria applied for mem- contributions to the social security sys- bership in the European Union, and tem and its spending, and the payment henceforth the OeNB was engaged in of wages, salaries and pensions. preparations for participation in the The cashless payment of wages, European Monetary System (EMS). salaries and pensions became popular One important task in this endeavor in Austria from the mid-1960s (Pohl, was the alignment of the Austrian pay- 1967), with the number of giro account ment system for large-value transac- holders augmenting accordingly. In 1970, tions with the corresponding system of credit institutions in Austria adminis- the upcoming Economic and Monetary tered 1.8 million giro accounts, in 1990 Union (EMU). 5.2 million and in 2015 10.7 million, The Deutsche Bundesbank began of which 5.3 million were dedicated to control and supervise the national to the payment of salaries and pen- payment organization earlier than the sions (Handler and Mooslechner, 1991, OeNB. The Bundesbank Act of 1957 pp. 95 and 100; OeNB, 2016). Accord- stipulates that the German central bank ing to Handler and Mooslechner (1991), “shall arrange for the execution of do- the increase in the number of giro ac- mestic and cross-border payments and counts as a percentage of the popula- shall contribute to the stability of pay- tion from 12% in 1963 to 67% in 1990 ment and clearing systems.” The fail- illustrates the trend toward a reduced ure of Herstatt Bank in 1974, the first importance of cash in the payment pro- bank insolvency to occur in the indus- cess (Handler and Mooslechner, 1991, trialized world in decades, laid open p. 99). In 2015, the number of giro ac- the risks connected with the settlement counts exceeded the Austrian population of foreign exchange transactions: If the by 25%, and accounts for the payment partners of a contract are trading in

13 OeNB Annual Reports 1956 to 1988 with the financial statements.

132 OESTERREICHISCHE NATIONALBANK Cashless payments in Austria: the role of the central bank

Chart 3 Turnover of OeNB giro accounts from 1956 to 1988 ATS billion Turnover of nominal GNP in % 16,000 12

14,000 10 12,000 8 10,000

8,000 6

6,000 4 4,000 2 2,000

0 0 1956 1958 1960 1962 1964 1966 1968 1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 Turnover of nominal GNP (right-hand scale) Turnover (left-hand scale)

Source: OeNB Annual Reports (1956–1988). different time zones, payments are services during the previous decades made at different points in time. In case had contributed to a closer integration of the illiquidity of one party, the coun- of financial markets worldwide. Analy- terparty perhaps does not get the quid ses commissioned by the EU central pro quo for its own payment. To reduce bank governors had shown that most the settlement risk, operational princi- national payment systems exhibited ples had to be changed to “payment risk factors which could endanger the versus payment” for foreign exchange stability of financial markets. In partic- transactions and to “delivery versus ular, the insolvency of one bank could payment” for securities trading (Deut- have major spillover effects because the sche Bundesbank, 2009). The payment settlement of payments of a large num- systems constructed by central banks in ber of national and international partic- the 1990s for the envisaged euro area ipants might not be finalized and cen- incorporated these maxims, comple- tral banks would have to provide emer- mented, inter alia, by the core principal gency liquidity. Therefore, central of finality. banks considered it their task to control every aspect of the payment process to 5 New duties for the OeNB in ensure efficient and secure monetary the payment process processes. The “lines of action” were In its Annual Report for the year 1992, laid down in the “Issues of Common the OeNB for the first time analyzed in Concern to EC Central Banks in the a separate chapter the importance of an Field of Payment Systems” and in the efficient payment system for the stabil- so-called Lamfalussy Standards.15 In ity of financial markets.14 The liberal- 1998 the Committee on Payment and ization and deregulation of financial Settlement Systems (CPSS) of the

14 OeNB Annual Report 1992, p. 58ff. 15 See Working Group on EC Payment Systems (1992); Working Group on EC Payment Systems (1993); Habacht (2001a, p. 128ff.).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 133 Cashless payments in Austria: the role of the central bank

Group of Ten (G-10) set up a task force Member States, domestic payment to define Core Principles for Systemi- transactions (mostly, but not only, cally Important Payment Systems (Bank small value transactions like the pay- for International Settlements, 2001), ment of rents, wages or utility bills) be- which were updated from time to time tween individual banks – were cleared to incorporate practical experiences and settled within the respective bank- gained. ing sector (savings banks, Raiffeisen In 1991, two years after Austria’s sector, and Volksbank credit coopera- application for EU membership, the tives, etc.) or on the basis of strict bilat- OeNB joined with a partner to set up a eral settlement procedures within the company to analyze the necessary adap- correspondent banking system. The tations of the Austrian payment system handling of domestic payment transac- for large-value transactions and to de- tions in Austria required 8,000 bilateral fine a strategy for the implementation account connections. of a system compatible with the one en- In 2011, the OeNB set up a clearing visaged for the euro area.16 In particu- service for Austrian domestic payment lar, such a system was to be a real time transactions (Clearing Service.Austria gross settlement (RTGS) system, i.e. – CS.A) within GELDSERVICE transactions should be booked immedi- AUSTRIA Logistik für Wertgestionierung ately on the accounts of the paying and und Transportkoordination G.m.b.H. receiving parties at their central bank (GSA) to simplify the processing of and should have finality. At the begin- transactions between financial institu- ning of the monetary union in 1999 the tions in different sectors and to make it various national RTGS systems of the easier for credit institutions to access euro area countries were integrated to national and international payment sys- form TARGET, the Trans-European tems. The settling of interbank pay- Automated Real-time Gross settlement ments at CS.A makes payments more Express Transfer system. From 1999, secure and efficient by reducing non- on, Austrian banks were connected to payment risks and lowering bank the euro area’s TARGET system via the liquidity requirements through the net- Austrian Real Time Interbank Settle- ting of interbank balances. Lower ment (ARTIS) system, which was re- liquidity requirements in turn reduce placed in 2007 by the Home Account- the need for equity capital and the ing Module Austria (HOAM.AT).17 In associated risk costs for the entire financial 2014, Austrian banks transferred 3 million marketplace. CS.A is a recognized sys- transactions with a total volume of EUR tem under the Settlement Finality Act. 23,674 billion via RTGS (HOAM.AT It is connected to the Eurosystem’s and TARGET2). TARGET2 system. In 2015, CS.A cleared Another recent innovation concerns a total of 570 million transactions with the clearing and settlement of domestic a total value of EUR 3 billion. 70% of payment transactions. Historically in these transactions were offset during Austria, unlike in all other 27 EU the clearing process; as a result only

16 This company was named STUZZA (Studiengesellschaft für Zusammenarbeit im Zahlungsverkehr – Association for Cooperation in Payment Transfers); the OeNB’s partner was GABE (Geldausgabeautomaten-Service Ges.m.b.H. – Company for Servicing Automated Teller Machines). 17 OeNB Annual Report 1998, p. 67ff; OeNB (2009, p. 64).

134 OESTERREICHISCHE NATIONALBANK Cashless payments in Austria: the role of the central bank

EUR 1 billion had to be actually trans- Table 3 ferred between participating banks. Transactions in payment and securities settlement systems The OeNB was obligated by a 2001 amendment to the Nationalbank Act18 2009 2014 to supervise payment systems in Number of Value in EUR Number of Value in EUR transactions in billion transactions in billion Austria. Hence, the central bank has million million to take measures to control the legal, RGTS 2 13,905 3 23,674 financial, organizational and technical Securities risks related to the service and use of settlement 2 365 2 377 payment systems. Retail payments 574 46 1,0051 721 International Modern payment systems can be payments 31 1,225 113 2,463 divided into four categories: Source: OeNB (2015, p. 106).

• large-value payment systems like 1 TARGET; 2013. • systems for the settlement of trans actions in securities; the euro area-wide RTGS. The security • retail payment systems, e.g. credit and efficiency of this system is of the cards, ATMs, POS terminals; greatest importance for the stability of • electronic money systems, “electronic the financial sector: The failure of the purses,” for the payment of very small settlement of a large money transfer amounts (Habacht, 2001b, p. 2). could have unwelcome knock-on effects. The OeNB discharges its duty to In 2013, more than one billion re- supervise payment systems in Austria tail transactions were executed by the through regular checks and on-site in- Austrian payment systems supervised spections on special occasions to check by the OeNB. The volume of EUR the organizational structure, the pro- 72 billion equals 41% of private con- cess flow and the technical status of the sumption expenditures of EUR 174 systems under supervision. In addition, billion, approximately the percentage companies that run payment systems of household purchases made by using have to supply the OeNB with qualita- bank or credit cards.19 A survey of the tive and quantitative data about their European Central Bank (ECB) shows operations. If such data were to show that Austria, along with Italy and Spain, peculiarities, the OeNB would imme- is still among those euro area coun- diately start a special examination of tries which use cash payments most the respective system. frequently, whereas France and the The number and volume of trans Netherlands use banknotes the least.20 actions of payment systems in operation According to the survey described in Austria is enormous, as can be seen in Mooslechner et al. (2012), it is not from table 3. By far the largest amounts the lack of cashless payment facilities are transferred via the Austrian leg of which is responsible for the high per-

18 Finanzmarktaufsichtsgesetz (Financial Market Supervision Act), Federal Law Gazette No. 97/2001, Article XIX). 19 See Mooslechner et al. (2012, p. 56). The difference of 6 percentage points between the payments made cashless according to the survey commissioned by the OeNB and the OeNB’s Financial Stability Report 30 (Table A25) can be explained by regular payments of households, e.g. for rents, electricity and gas, which are usually made by standing orders. 20 See European Central Bank (2011, p. 79). This article gives no explanations for the widely differing preferences for using cash.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 135 Cashless payments in Austria: the role of the central bank

centage of cash payments in Austria but mainly agricultural activities, to the primarily consumers’ preferences. abundance of paper money as a conse- Cash apparently best fulfills the de- quence of the financing of major military sired requirements of a medium of pay- campaigns through central bank loans ment: paying with cash is quick and and the issuing of paper money by the easy and entails no additional costs, and state. Until the last quarter of the 19th the many ATMs facilitate the sourcing century a cash-saving payment infra- of cash (Mooslechner et al., 2012, structure was not urgently needed in p. 75f).21 Although the use of other pay- the territory of the Habsburg monarchy. ment media has increased lately, there The incorporation of the Austrian is still a persistence of behavior par- Postal Savings Bank in 1882 marked a ticularly among older people. The in- turning point in the spreading of mod- creasing volume of goods purchased ern financial technologies in Austria. from internet suppliers did not alter The more than 4,000 post offices of- the structure of the payment process fered giro and savings accounts for a significantly: in 2011 a mere 0.13% of stratum of the population that previ- all recorded transactions – 0.24% vol- ously had no contact to banking. An ume-wise – were settled using pay- amendment to the charter of the ment media dedicated to internet Austro-Hungarian Bank in 1887 was a transactions (e.g. PayPal, Clickand- precondition for the substantial growth Buy) (Mooslechner et al., 2012, p. 65). in the number of giro accounts and These observations suggest that cash transaction volumes of the central bank will dominate the retail payment pro- in the following years. Giro accounts at cess in Austria for many years to come. the central bank were used mainly for It will be a very important task of the clearing of balances resulting from central banks in the future to supervise commercial and financial transactions payment systems in order to minimize of the banking community. Saving above all the technical and operational banks and cooperative banks did not risks. Hence, central banks as supervi- offer giro accounts and were integrated sors have to take care that the integrity into the economy-wide payment struc- of efficient and complex payment sys- ture only after WW II. tems is preserved by aiming to have the In the interwar period, Viennese most technically advanced systems in and Austrian banks had to adapt to the operation. economic and financial potential of a far smaller economic area. This happened 6 Conclusions not as a planned process but as a result Austria was a latecomer in the use of of severe and costly banking crises in cash-saving or cashless payment tech- the 1920s and 1930s. The transaction niques. This can be attributed to the eco- volumes in the payment system, in par- nomic backwardness of the Habsburg ticular financial transactions on the monarchy compared with the North- stock exchange, were severely hit by western European countries in the first the Great Depression. The economic decades of the 19th century, to the low climate in the period between WW I degree of monetization of economic, and WW II was not conducive to the in-

21 Tax evasion could also be an argument for using cash, but this would further increase the high percentage of cash payments, although probably not significantly.

136 OESTERREICHISCHE NATIONALBANK Cashless payments in Austria: the role of the central bank

troduction or spreading of new financial system, which had to be aligned with technologies. the system in place in the euro area. The economic boom in the Second For large-value transactions, Austrian Austrian Republic and the rapid in- banks could use ARTIS, later HOAM. crease of production, employment and AT, to transact via TARGET with the income resulted in a wider use of bank- ECB or with other euro area banks. The ing services. The number of giro and turnover of retail payment systems in savings accounts grew strongly, in part Austria is steadily increasing, although because since the mid-1960s salaries, the majority of household purchases are wages and pensions were paid by using still paid by cash. The OeNB’s duties banking facilities. There were several as regulator and supervisor of payment payment system networks in place cor- systems in Austria were laid down in an responding to the different banking sec- amendment to the Nationalbank Act in tors. The OeNB acted primarily as the 2001. Before that, the central bank was central clearing house. The spread of aware of the importance of a well- electronically based systems in banking functioning, safe payment system for changed the payment process as the use financial stability but did not have a of automated teller machines, point-of- strict legal obligation to ensure it. sale terminals and credit cards increased. Preserving the integrity of payment The introduction of the euro in systems will be one of the main challenges Austria in 1999 had a basal influence on in the future. the structure of the Austrian payment

References Bank for International Settlements. 2001. Committee on Payment and Settlement Systems. Core Principles for Systemically Important Payment Systems. Basel. Born, K. E. 1993. Geld und Währung im 19. Jahrhundert. In: Pohl, H. (ed.). Europäische Banken- geschichte. Frankfurt am Main. Bubenik, F. 1888. Die Technik des Giro-Verkehrs bei der Oesterreichisch-ungarischen Bank. Vienna. Butschek, F. 1999. Statistische Reihen zur österreichischen Wirtschaftsgeschichte. Vienna. Deutsche Bundesbank. 2009. Cashless payments in Germany and the role of the Deutsche Bundesbank. In: Monthly report. March. 49–64. Ebner, H. 1962. Der Giroverkehr in Österreich. Vienna. European Central Bank. 2011. The use of euro banknotes – results of two surveys among households and firms. In: Monthly Bulletin. Finanzmarktaufsichtsgesetz (Financial Market Supervision Act). www.ris.bka.gv.at/ GeltendeFassung.wxe?Abfrage=Bundesnormen&Gesetzesnummer=20001456 (retrieved on August 2, 2016). Fischer, P. D. 1893. Postsparkassen. In: Conrad, J., W. Lexis, L. Elster und E. Loening (eds.). Handwörterbuch der Staatswissenschaften 5. Jena. 218–225. Gesetz über die Deutsche Bundesbank (Bundesbank Act). www.bundesbank.de/Reda- ktion/EN/Downloads/Bundesbank/Organisation/bundesbank_act.pdf?__blob=publicationFile (retrieved on August 2, 2016). Giefing, A. and W. Auracher. 1977. Entstehung und Entwicklung der Zentralinstitute im Sparkassensektor. In: Österreichisches Forschungsinstitut für Sparkassenwesen (ed.). Funk- tionen der Girozentrale im Kreditapparat. Vierteljahres-Schriftenreihe 4/77. 7–16.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 137 Cashless payments in Austria: the role of the central bank

Girozentrale der österreichischen Genossenschaften. 1929. Rechenschaftsbericht über das Geschäftsjahr 1928. Vienna. Goodhart, C. 1988. The Evolution of Central Banks. Cambridge, Mass., and London. Gross, N. 1983. Austria-Hungary in the World Economy. In: Komlos, J. (ed.). Economic Develop- ment in the Habsburg Monarchy in the Nineteenth Century. New York. 1–45. Governing Board of the OeNB. Various years. Mitteilungen des Direktoriums der OeNB. Vienna. Habacht, R. 2001a. The Future Role of the OeNB as the Body Responsible for Payment Systems Oversight in Austria. In: Financial Stability Report 2. OeNB. 121–139. Habacht, R. 2001b. Aufgaben der Zahlungssystemaufsicht. Themenblätter Wirtschaft & Finanzen 11. OeNB. Hammerschlag, P. 1905. Abrechnungsstellen. In: Mischler, E. and J. Ulbrich (eds.). Österreichisches Staatswörterbuch. Second edition 1. 18–23. Handler, H. and P. Mooslechner. 1991. Zahlungsverkehrssysteme und Zahlungsverkehr in Österreich. Austrian Institute of Economic Research. Vienna. Hauptverband der österreichischen Sparkassen (ed.). 1972. 150 Jahre Sparkassen in Österreich 1. Vienna. Houtman-De Smedt, H. and H. van der Wee. 1993. Die Entstehung des modernen Geld- und Finanzwesens Europas in der Neuzeit. In: Pohl, H. (ed.). Europäische Bankengeschichte. Frankfurt am Main. 75–120. Jobst, C. and H. Kernbauer. 2016. The Quest for Stable Money. Central Banking in Austria, 1816–2016. Campus. Frankfurt am Main and New York. Kahn, C., S. Quinn and W. Roberds. 2014. Central Banks and Payment Systems: The Evolving Trade-off between Cost and Risk. Paper in preparation for the Norges Bank conference “Of the Uses of Central Banks: Lessons from History.” Oslo. June 5–6. Kernbauer, H. 1991. Währungspolitik in der Zwischenkriegszeit. Geschichte der Oester reichischen Nationalbank von 1923 bis 1938. Vienna. Klauhs, H. 1977. Bilanz über 50 Jahre Genossenschaftliche Zentralbank AG. In: Genossenschaft liche Zentralbank AG (ed.). 50 Jahre Genossenschaftliche Zentralbank Aktiengesellschaft 1927– 1977. Vienna. 7–11. k. k. Finanzministerium. 1895. Mitteliungen des k.k. Finanzministeriums. Vienna. Koch, R. 1910. Giroverkehr. In: Conrad, J., W. Lexis, L. Elster und E. Loening (eds.). Handwörter- buch der Staatswissenschaften. Third edition (5). Jena. 14–30. Komlos, J. 1983. The Diffusion of Financial Technology into the Austro-Hungarian Monarchy toward the End of the Nineteenth Century. In: Komlos, J. (ed.). Economic Development in the Habsburg Monarchy in the Nineteenth Century. New York. 137–164. Körner, M. 1993. Handel und Geldwesen im mittelalterlichen Europa. In: Pohl, H. (ed.). Euro päische Bankengeschichte. Frankfurt am Main. 33–72. März, E. und K. Socher. 1973. Währung und Banken in Cisleithanien. In: Brusatti, A. (ed.). Die Habsburgermonarchie 1848 – 1918 1. Die wirtschaftliche Entwicklung. Vienna. 323–368. Matis, H. 2005. Die Schwarzenberg-Bank. Kapitalbildung und Industriefinanzierung in den habsburgischen Erblanden. Vienna. Matis, H. und F. Weber. 1993. Kaisertum Österreich – Donaumonarchie. In: Pohl, H. (ed.). Europäische Bankengeschichte. Frankfurt am Main. 316–332. Mecenseffy, E. 1896. Die Verwaltung der Oesterreichisch-ungarischen Bank 1886–1895. Vienna. Mises, L. 1924. Theorie des Geldes und der Umlaufsmittel. Vienna and Leipzig. Mooslechner, P., H. Stix and K. Wagner. 2012. The Use of Payment Instruments in Austria. A Study Based on Survey Data from 1996 to 2011. In: Monetary Policy and the Economy Q4/12. OeNB. Vienna. 53–77.

138 OESTERREICHISCHE NATIONALBANK Cashless payments in Austria: the role of the central bank

OeNB. 2009. Zahlungsverkehrsbericht 2009. Vienna. OeNB. 2015. Financial Stability Report 30. Vienna. OeNB. 2016. Number of Deposit Accounts – all banking sectors. www.oenb.at/isaweb/report. do?report=3.8 (retrieved on June 27, 2016). OeNB. Various years. Annual Report. Vienna. Oesterreichisch-ungarische Bank. 1887–1913. Jahressitzung der Generalversammlung der Oesterreichisch-ungarischen Bank. Vienna. Österreichisches Forschungsinstitut für Sparkassenwesen (ed.). 1961. Bargeldlose Lohn- und Gehaltszahlung. Schriftenreihe des Österreichischen Forschungsinstituts für Spar- kassenwesen 1(1/2). Vienna. Peter, R. (n.d.). Entwicklung des konventionellen Zahlungsverkehrs. In: Österreichisches For- schungsinstitut für Sparkassenwesen (ed.). Zahlungsverkehr im Umbruch. Schriftenreihe des österreichischen Forschungsinstitutes für Sparkassenwesen 16(2). Pohl, E. 1967. Die bargeldlose Entlohnung. In: Österreichisches Forschungsinstitut für Sparkassen- wesen (ed.). Schriftenreihe des österreichischen Forschungsinstitutes für Sparkassenwesen 7(4). Pressburger, S. 1959. Das österreichische Noteninstitut 1816–1966 I/1. Vienna. Pressburger, S. 1962. Das österreichische Noteninstitut 1816–1966. I/3. Vienna. Quinn, S. and W. Roberds. 2006. An Economic Explanation of the Early Bank of Amsterdam, Debasement, Bills of Exchange, and the Emergence of the First Central Bank. Federal Reserve Bank of Atlanta. Working Paper 2006-13. Rauchberg, H. 1886. Der Clearing- und Giro-Verkehr. Ein statistischer Beitrag zur Kenntnis des volkswirtschaftlichen Zahlungsprocesses. Vienna. Rauchberg, H. 1897. Der Clearing- und Giro-Verkehr in Österreich-Ungarn und im Auslande. Vienna. Sadleder, W. 1977. Die Bedeutung der Girozentrale für die Sparkassen. In: Österreichisches Forschungsinstitut für Sparkassenwesen (ed.). Funktionen der Girozentrale im Kreditapparat. Vierteljahres-Schriftenreihe 4/77. 23–30. Szedenik, A. F. 1937. Die Entwicklung des bargeldlosen Zahlungsverkehrs unter besonderer Berücksichtigung Österreichs. Dissertation. Hochschule für Welthandel. Vienna. Teichova, A. 1997. Banking and industry in central Europe, nineteenth to twentieth century, In: Teichova, A., G. Kurgan-Van Hentenryk and D. Ziegler (eds.). Banking, trade and industry. Europe, America and Asia from the thirteenth to the twentieth century. Cambridge. 214–228. von Lederer, C. 1847. Die privilegirte österreichische National-Bank. Ihre Gründung, ihre Entwicklung und ihr Wirken. Vienna. Wagner, M. and P. Tomanek. 1983. Bankiers und Beamte. Hundert Jahre österreichische Postsparkasse. Vienna. Wärmer, G. 1936. Das österreichische Kreditwesen. Vienna. Wood, J. H. 2011. A History of Central Banking in Great Britain and the United States. New York. Working Group on EC Payment Systems. 1992. Report to the Committee of Governors of the Central Banks of the Member States of the European Economic Community. Issues of Common Concern to EC Central Banks in the Field of Payment Systems. Basel. Working Group on EC Payment Systems. 1993. Report to the Committee of Governors of the Central Banks of the Member States of the European Economic Community. Minimum Common Features for Domestic Payment Systems. Basel. Zuckerkandl, R. 1899. Oesterreichisch-ungarische Bank. In: Conrad, J., Elster, L., Lexis, W. and E. Loenig (eds.). Handwörterbuch der Staatswissenschaften 2. Second edition. Jena. 222–256.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 139 Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

Clemens Jobst, In this study, we provide a discussion of the role the Austrian central bank played as a lender Kilian Rieder1 of last resort (LLR) during selected episodes of financial distress from the Nationalbank’s foundation in 1816 until the Creditanstalt crisis of 1931. Based on our evidence, we argue that free lending as advocated by British economist Walter Bagehot was historically the exception rather than the rule in Austria, and that no clear evolution toward more “free lending” is observable over time. The panic of 1912, a particularly fascinating example of a “forgotten” crisis that has never been investigated in detail, serves as our benchmark because the Natio nalbank’s crisis management during this specific episode comes very close to an effective case of free lending. Instances of credit rationing during other financial crises seem to have emerged as a consequence of public doubts about the value-storing capacity of banknotes and due to a lack of discountable or pledgeable assets resulting from the Nationalbank’s regulations and/ or risk management framework. Our study echoes earlier literature in the field, underlining the importance of the microeconomics of last resort lending, including the incentive structure of lending programs and the ex ante supervision of counterparties. JEL classification: E58, G01, N13, N14 Keywords: central bank, Austria, Oesterreichische Nationalbank, lender of last resort, financial crisis, banking crisis, credit rationing, liquidity crisis, bank run, moral hazard, Bagehot

Why have central banks’ responses to academics, ever since Henry Thorn- financial crises differed so much over ton’s 1802 treatise on The Paper Credit of time? One explanation might simply Great Britain (Thornton, 1802). In its be endogenous learning: policymakers simplest specification, this theoretical draw lessons from mistakes committed and empirical discussion turns around in the past and adapt crisis manage the conditions and circumstances under ment accordingly. Undoubtedly, eco- which central banks should provide an nomic (history) research on the Great extra liquidity injection into the finan- Depression of the 1930s has informed cial system for the benefit of all banks monetary policy reactions to the recent under circumstances of a collective finan- financial turmoil of 2008 (Almunia et cial market liquidity crisis.2 al., 2010; Eichengreen, 2015). How- Walter Bagehot is famously taken to ever, a focus on lessons from history have answered this question in the fol- neglects the fact that the rationale and lowing way: Central banks should lend impact of central banks’ responses to freely, at high interest rates, and only in crises have always been deeply con- return for good collateral.3 The myriad tested, both by contemporaries and rationalizations of Bagehot’s principles

1 Oesterreichische Nationalbank, Economic Analysis Division, [email protected]; University of Oxford, Department of Economics and History Faculty, [email protected]. The views expressed in this paper are exclusively those of the authors and do not necessarily reflect those of the OeNB or the Eurosystem. The authors would like to thank Vincent Bignon (Banque de France) and Stefano Ugolini (Université de Toulouse) as well as the referee and the participants in a workshop among the authors of this volume for very helpful comments and valuable suggestions. Walter Antonowicz and Claudia Köpf from the OeNB’s Bank History Archives have greatly Refereed by: facilitated access to primary sources. Daniel Maier provided excellent research assistance. Charles Goodhart, 2 Rather than representing emergency liquidity assistance (ELA) to individual banks, the historical concept of last London School of resort lending was thought of as a form of support extended to the financial system as a whole. See Goodhart (1999). Economics 3 These principles are generally derived from various passages in Bagehot’s major work. See Bagehot (1873).

140 OESTERREICHISCHE NATIONALBANK Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

gave rise to many myths and misunder- ture to advance our take on Bagehot’s standings about the lender of last resort principles. Section 2 provides a brief (LLR) by diverting attention from the sketch of the monetary policy frame- historical context in which these func- work of the Nationalbank during the tions first emerged.4 In this study, we period under study. Section 3 then aim at retracing these roots in a parti describes the “forgotten” panic of 1912, cular national setting, thereby adding which we argue to be a benchmark case Austria to the existing international of Bagehot-style free lending. Section 4 literature on LLRs, which is almost ex- looks at why free lending, which worked clusively dominated by studies on the so successfully in 1912, was not adopted U.S.A. and the U.K.5 We analyze the during most other financial crises. Sec- Nationalbank’s behavior in the context tion 5 concludes. of six crises – 1820, 1848–49, 1873, 1912, 1923–24, and 1931. Our selec- 1 The central bank as lender of tion encompasses both well-known and last resort less well-known episodes of financial Financial crises have generated an ex- distress, but does not represent a com- tensive literature in both economics and prehensive list; rather, it reflects our economic history (Allen et al., 2009). attempt to capture the manifold faces While the proximate and ultimate and aspects of LLR activities through- causes of the fragility leading to a finan- out history. While the crisis of 1820 cial crisis may differ and continue to be constitutes the first financial panic debated, all financial crises share some following the foundation of the privilegirte commonalities that are most relevant oesterreichische National-Bank in 1816, for the question of last resort lending the crash of 1931 in turn stands out as a (Gorton, 2012). These shall be briefly natural endpoint: it represents the last outlined here. crisis before financial repression in the aftermath of World War II (WW II) 1.1 The logic of financial crises led to the complete disappearance of Market economies depend on the use as banking crises in all industrial coun- transaction media of short-term debt tries until the 1970s. obligations issued by financial inter Based on our evidence, we argue mediaries. These obligations can take that in Austria, free lending was histori- many forms. Normally, we think of cally the exception rather than the rule. deposits here but historically, banks The reason was not that the Nationalbank have also issued private notes or, more did not care for financial stability; recently, shares in money market funds rather, its policy was constrained by a or repos. These instruments have in range of factors, including regulation, common that their exchange requires limits to information and confidence in little or no information on the issuer or the Nationalbank itself. The remainder the underlying collateral; in normal of this study is structured as follows: times, they are information insensi- We first briefly review the LLR litera- tive.6 Banking crises are characterized

4 See Goodhart (1999) and Bignon et al. (2012) for a more in-depth discussion of these fault lines. 5 Some international perspectives are provided in Kindleberger and Aliber (2011). For in-depth studies of individual countries, see e.g. Buyst and Maes (2008) on 19th-century Belgium, and Bignon et al. (2012) on France. 6 Information-insensitive assets are safe assets in the sense that they are accepted as collateral without fear of adverse selection and can store value over time. See Gorton and Ordoñez (2014) and Gorton (2016).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 141 Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

by a flight out of these debt obligations to withdraw cash. However, panics may and into cash. They are systemic events also affect wholesale funding, as during in the sense that they involve many or the recent financial crisis of 2008 most financial intermediaries (Gorton, (Gorton and Metrick, 2012). Ultimately, 2012). panics might even target the liabilities Runs constitute a rational response issued by the central bank itself, nor- to shocks that cause creditors to doubt mally the most liquid and safe asset the ability of financial institutions to available, when people start to doubt honor their debt contracts. These shocks the future value of banknotes and strive may be very small in nature and might to exchange banknotes into real or for- even be expected to happen only in the eign assets. Despite the different ap- future. Yet, it suffices that they are pearance of panics, the consequences strong enough to induce creditors to are similar. In a systemic run, the question the quality of their claims, affected intermediaries are by defini- thereby turning previously information unable to fully meet the large-scale tion-insensitive into information-sensi- scramble for cash they are facing. In tive assets. When uncertainty about this sense, all intermediaries become asset values increases, the debt capacity insolvent, as they cannot honor their of collateral – the amount of secured obligations without trying to sell assets borrowing that can be sustained by an (Gorton, 2012). If many credit institu- asset – can fall dramatically (Acharya et tions sell their assets simultaneously, al., 2012). Creditors will be inclined to however, fire sales will result, as asset ration their funds by converting their values are depressed by “cash-in-the- debt into safer and more liquid assets or market pricing.”7 The resulting price by increasing the haircuts on short-term declines can have severe macroeco- collateralized debt. If shocks effectively nomic consequences and may affect trigger the production of information parts of the banking sector, other third about the quality of underlying assets, parties and, more globally, the real debt backed by collateral that is revealed economy, all of which had not been to be bad may be called in outright related to the initial run (Antinolfi et (Gorton and Ordoñez, 2014). Whether al., 2015; Brunnermeier et al., 2009). newly available information about col- These negative externalities provide lateral quality sparks a systemic run the rationale for a LLR that can provide rather than being merely translated into the financial system with unlimited increasing margins eventually depends cash in exchange for the liabilities tar- on a variety of additional factors, such geted by the run. In doing so, the LLR as the scale of the information produc- can address creditors’ doubts, stop the tion, the length of the preceding credit incipient run and avert fire sales that boom and the economic as well legal would have potentially severe macroeco- microstructure of short-term secured nomic consequences (Oehmke, 2014). funding markets (Martin et al., 2014). In fact, the mere setup of LLR might Systemic financial crises can take prevent the panic altogether. To be able different forms. The most emblematic to fulfill its task, the LLR must be an example is a depositor run, when savers institution that is in a position to issue physically queue in front of their banks an unlimited (or at least a very large)

7 “Cash-in-the-market pricing” occurs when the total liquidity available in the market at a specific moment in time is (much) smaller than the total value of assets offered for sale. See Allen and Gale (2007).

142 OESTERREICHISCHE NATIONALBANK Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

amount of cash, in other words, an insti- be dear, still money is to be had” (Bagehot, tution that cannot become illiquid itself. 1873). Bignon et al. (2012) propose a Historically, in the absence of central simple test for the absence of credit banks, associations of large banks have rationing. Under the null hypothesis of served as LLRs (see e.g. Gorton, 2012). free lending, the central bank’s “inter- Another candidate is the government, est rate (for any given quality) ought to which has the advantage of being able always be above, or equal to, the market to tax. As monopoly issuers of cur- rate (for the same quality)” (Bignon et rency, however, central banks are in al., 2012). The market rate can only be the most natural position to act as higher than the policy rate in the pres- LLRs. ence of credit rationing; under free lending, an initially higher market rate 1.2 Bagehot’s principles would decrease immediately due to What exactly is the LLR’s task? On the funding liquidity arbitrage. surface, each financial crisis looks new Naturally, this test immediately raises and different. It involves new financial the question of how high the quality products and possibly new agents.8 At level of the paper is to which the policy the same time, the basic mechanism rate applies. If this quality level is set underlying crises is always the same – very high, credit rationing might effec- the flight out of debt into cash – and tively result from counterparties’ short- hence allows for the development of age of eligible collateral. The crucial some general guidelines. The most question is then which assets the central famous guidelines come from Bagehot’s bank perceives as “good collateral.” book Lombard Street (Bagehot, 1873) Bagehot (1873) himself defined “good and were condensed by later authors collateral” as financial paper which “in into three principles, namely that cen- ordinary times is reckoned a good secu- tral banks should lend freely (i.e. with- rity” and is “commonly pledged and out limits), at high interest rates and easily convertible.” In ordinary times, only in return for good collateral.9 however, central banks may have many Debates on the significance and reasons to limit the type or quality of meaning of these three “rules” continue assets admissible for their credit opera- today. Yet they can serve as a useful tions. Broader lists of eligible securities framework for organizing the following (in the case of lombard loans) or less discussion of the Nationalbank’s policy stringent criteria in discounting will be during financial crises, particularly so associated with heightened credit risk as Bagehot’s work is grounded in the or at the very least will increase opera- practice of 19th-century central banking. tional expenses. As a result of these Bagehot’s “free lending” principle is restrictive practices ex ante, counter- best read as a plea against credit parties may become short of eligible rationing by the central bank. To halt assets when a crisis occurs, thus pre- bank runs, the LLR needs to propagate venting the central bank from providing “the impression that though money may the required liquidity ex post (Bindseil,

8 As historical experience shows (most recently the panic of 2008), crises are bound to happen in those areas of the financial system that are not regulated or guaranteed, which is probably also the main reason why deposit insurance, an alternative means to keeping up confidence in bank liabilities, is not sufficient. In Austria, deposit insurance was only introduced in 1979, and is thus of no relevance for the period under study. 9 As already pointed out above, these principles never appear as such but are generally derived from various passages in Bagehot (1873).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 143 Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

2014). No matter how this tradeoff is (2009), Bagehot mainly cared about the solved in the concrete case, free lend- efficient use of limited central bank re- ing requires that the LLR at least does serves in the context of a commodity not tighten eligibility criteria (for dis- money regime. Although the Bank of counts) and/or reduce the pool of eligi- England’s requirement to abide by pre- ble securities (for lombard loans) in defined cover requirements was usually response to financial distress. Inertia suspended in case of a financial panic, should prevail in accepting bills and the credibility of the Bank of England assets that are considered “good collat- critically hinged on prudent manage- eral” in normal times (Bindseil, 2014). ment of available reserves. Too large an While the logic of “free lending” and expansion of the money supply over “good collateral” is relatively straight- and above the statutory limit, risked forward, the concept of “high rates” has triggering a currency crisis. High rates been found to be more difficult to ra- thus served to efficiently allocate lim- tionalize. One standard interpretation ited reserves among the counterparties relates to moral hazard. High rates pro- that needed them most. The argument vide an incentive for banks to engage in advanced by Bignon et al. (2012) is proper liquidity risk management in more akin to the current justification of normal times. However, this rational- negative interest rates on central bank ization has been challenged along sev- deposits. It rests on the observation eral lines. On the one hand, high inter- that during banking panics, financial est rates might just encourage failed in- intermediaries search for a safe haven stitutions to gamble for survival. On (central bank deposits), which curtails the other hand, recent theoretical work the supply of funds on the money mar- shows that moral hazard might be less ket. In the extreme case, the central of an issue as long as informational bank would internalize the entire asymmetries on counterparty risk and/ money market, lending to banks in or collateral quality between borrowers need of funds while taking in deposits and the central bank are not large from banks with surplus funds. In addi- (Martin, 2006; Castiglionese and Wag- tion to straining the limited reserves of ner, 2012; Naqvi, 2015). In fact, cen- the central bank as discussed above, tral banks have never stayed at arm’s such an approach would mean that in- length from their borrowers, not in the termediaries also forgo the use of their past and much less today. The intensifi- private knowledge on counterparties as cation of counterparty management well as collateral and stop collecting in- and banking supervision is indeed formation, which might hamper the strongly correlated with central banks’ functioning of the market in the long assumption of LLR responsibilities. For run. With interest rates on central bank example, Flandreau and Ugolini (2014) deposits always at zero in the 19th cen- point to the crucial importance of con- tury, the main lever to encourage banks tinuous ex ante monitoring of counter- to transact with each other is an in- parties for the success of last resort crease in the discount rate, which, in lending in England after the 1866 Bagehot’s words, will “operate as a heavy Overend crisis. fine on unreasonable timidity, and will Contributing to this debate, Martin prevent the greatest number of applica- (2009) and Bignon et al. (2012) have tions by persons who do not require it” proposed two alternative rationales for (Bagehot, 1873, p. 197). “high rates.” According to Martin

144 OESTERREICHISCHE NATIONALBANK Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

2 The operational framework of collateral, typically securities, at a set the Nationalbank interest rate, the so-called lombard To analyze how a central bank provides rate. The conditions were straightfor- liquidity during a crisis, it is useful to ward. Loans had a maximum maturity understand the design of its normal of three months. The maximum amount -time operational framework first. Here, of the loan was calculated as the market the central bank’s standing facilities price of the pledged securities minus a play an important role. Standing haircut. If the price of the collateral- facilities are those operations whose ized securities dropped below a speci- general lending conditions (most im- fied threshold during the term, the portantly the eligibility of assets and debtors had to provide additional col- the applicable interest rate) are defined lateral or repay the loan. The list of by the central bank, while actual oper- eligible securities and haircuts was pub- ations are initiated by the central bank’s lished.11 At the outset, it included only counterparties, not the central bank government bonds, but it was extended itself. By design, standing facilities can over time to encompass a wide range of accommodate sudden increases in subsovereign, railroad and covered liquidity demand without requiring the bonds as well as shares of selected rail- central bank to take any explicit mea- roads and shipping companies.12 Hair- sure. The only prerequisites are that cuts were set at relatively high levels the central bank does not restrict access compared to today, mostly at 25% or to the standing facility during a crisis 30% of the market value. and that the facility does not suffer In discount operations, on the other from a stigma, meaning that counter- hand, the Nationalbank bought a bill of parties feel it is safe to access the facil- exchange with a short initial or remaining ity if they need it (Bindseil, 2014). maturity at a discount to its nominal In Austria, liquidity-providing oper value (the discount rate).13 In the period ations were long dominated by dis- under study, access to discounting was counts and advances, as was the case for not limited to financial intermediaries: most other central banks in the 19th and the Nationalbank would accept bills for first half of the 20th century.10 Both discounting from “any man, regardless types of operations were organized as of social standing, if he is known to the standing facilities in principle, where Nationalbank as law-abiding.”14 Access rules and conditions were fixed in was, however, restricted geographi- advance and made publicly known. The cally, as the Nationalbank had to cash Nationalbank operated both facilities the bill at maturity and could therefore from its foundation in 1816 and kept only accept bills payable at locations at the following basic features unchanged which it had an office. Until 1841, throughout the period under study: when the first Nationalbank branch In an advance, the Nationalbank office was opened in Prague, only bills granted a loan against some pledged payable in Vienna were eligible for

10 On Austria, see Jobst and Kernbauer (2016). For an international comparison, see Jobst and Ugolini (2016). 11 This disclosure policy is in contrast to the Bank of England’s, where such information was not made public (Flandreau and Ugolini, 2014). 12 For lists of eligible central bank collateral, see Compass (1868–1919). 13 In principle, the Nationalbank also discounted treasury bills and short-dated securities like coupons, but their importance was negligible in practice. 14 Article 65 of the Reglement (a supplement to the Nationalbank’s statutes), see Pressburger (1959a).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 145 Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

discounting. In the following decades, to refuse an operation, e.g. the granting the geographical scope of eligibility was of an advance or the discounting of widened significantly, as the National- paper, without giving reasons. Thus, bank opened additional branch offices whether the Nationalbank effectively throughout the entire Habsburg mon- pursued a policy of free lending or not archy. Unlike payment modalities for cannot be judged on the rules set down advances, where the Nationalbank could in the operational framework alone. resort to selling the collateral posted in Rather, the question is an empirical one the case of nonpayment, the reimburse- that we will discuss in the sections that ment of a bill depended on the ability of follow. the acceptor or the drawer to pay at maturity. Assessing the quality of the 3 A textbook case of free lending: signatures featured on a bill was thus The “forgotten” panic of 1912 essential. To be eligible for discount, a In the following section, we use the bill as a rule had to be signed by three, panic of 1912 as our benchmark case. but in any case by at least two, “persons Apart from our finding that the known to be solvent.” The quality of a Nationalbank’s crisis management during bill was established in two stages. First, this specific episode comes very close the Nationalbank drew up lists of per- to an effective case of free lending, the sons and companies who were deemed panic of 1912 constitutes a particularly eligible in principle and defined a max- enticing case as a “forgotten” crisis that imum amount up to which the counter- is mentioned in historiography but that party could access the Nationalbank di- has never been investigated in detail rectly (by submitting a bill for discount) (Michel, 1976; Pressburger, 1973). or indirectly (by figuring as an acceptor The banking panic that gripped on a bill submitted by a third counter- Austria in November 1912 has to be party).15 The required information was seen in the context of the political ten- collected by the Nationalbank’s local sions following the defeat of the Otto- branch offices. Each branch hosted a man Empire in the First Balkan War standing discount committee com- (October 1912 to May 1913). The prob- posed of bank officials as well as local ability of a military involvement of businessmen and dignitaries. Based on Austria-Hungary peaked in the last two public and private information, these months of 1912 when Serbia’s ambi- committees regularly revised the indi- tions to annex Albanian territory met vidual credit limits (Kövér, 2015). Sec- with heavy opposition from the ond, the discount committee examined Habsburg empire and culminated in and checked each bill presented for dis- mutual threats of war. The tensions counting against the formal submission only eased when the Treaty of London criteria. was concluded in early 1913. The im- While the Nationalbank made these minent danger of an armed conflict in conditions public and while in principle fall 1912 served as an exogenous shock anybody who fulfilled the criteria could to Austro-Hungarian depositors with apply for a discount or an advance, the financial intermediaries in the border Nationalbank’s statutes and regulations regions with Serbia (Carniola, Croatia- left some scope for discretion. In par- Slavonia, Dalmatia, the Austrian Litto- ticular, they allowed the Nationalbank ral and Southern Hungary) and Serbia’s

15 In addition, the Nationalbank operated with aggregate limits at the branch level.

146 OESTERREICHISCHE NATIONALBANK Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

major ally, Russia (Galicia, Bukovina). turned increasingly cautious.17 The The preference for liquidity increased, business model of savings banks was fears were rife that the government particularly prone to maturity mis- might confiscate savings deposits in the matches, as such banks financed long- case of war, and uncertainty arose term mortgages with sight deposits. about (local) asset values in the event of Hence, the banks targeted by runs a hostile assault or an occupation. Bank found it difficult to obtain funding and creditors started to panic and with- risked illiquidity-induced defaults. drew a significant fraction of their sight The Austro-Hungarian Bank (Oester liabilities from credit institutions and reichisch-ungarische Bank) reacted by converted them into cash (chart 1). letting its standing facilities operate From early October, when the first freely and distributing liquidity gener- runs started, to the end of December ously. There are no signs that the bank 1912, banknotes in circulation increased restricted access to its discount window. from 2.4 billion crowns to 2.8 billion Over the entire year 1912, the bank crowns, or by some 15%.16 At the same rejected only some 63,000 bills for for- time, interbank credit was drying up, mal or other reasons out of a total of as large correspondent banks in the 4.6 million bills or, put differently, core cities, which themselves faced accepted 98.6% of all bills submitted tight conditions in money markets, for discounting. This percentage was

Chart 1 The forgotten panic of 1912 Change in savings bank deposits and central bank advances No credit rationing Million crowns % 50 7

40 6

30 5 20 4 10 3 0 2 –10

–20 1

–30 0 July Aug. Sep. Oct. Nov. Dec. Jan. Feb. Mar. April May June Istria Tyrol

Styria 1912 1913 Silesia Galicia Trieste Gorizia Moravia Salzburg Carniola Bohemia Dalmatia Bukovina Carinthia Vorarlberg Lower Austria Lower Upper Austria Change in savings banks deposits between December 1911 Oﬃcial discount rate set by the Nationalbank and December 1912 Market rate for prime bills Vienna Change in advances at central bank branches between December 1911 and December 1912

Source: OeNB (1913a), k. k. Statistische Zentralkommission (1915), Neue Freie Presse (1864–1939).

16 December typically also marked a seasonal peak in currency circulation. Still, the increase relative to the previous year topped out at 305 million crowns on December 23 and amounted to 275 million crowns on December 31, 1912. 17 E.g. Prague banks withdrew credit lines from Galicia. See OeNB (1912b, November 19).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 147 Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

somewhat lower than in 1911, when financial intermediaries in the affected 99.1% of bills were accepted, but above regions borrowed extensively on the the average rate of 97.6% that was security of covered bank bonds (1911: observed in the first five years of the 27%, 1912: 41%). 1900s. Moreover, the hypothesis of Besides free lending, the policy of free lending is also supported by the the Austro-Hungarian Bank contained Nationalbank’s handling of individual a second important aspect, namely its discount limits during the crisis. When effort to revive the interbank market. these limits became binding for a During the autumn of 1912, the number of credit institutions in late Nationalbank raised its policy rates 1912, the bank relaxed the limits with- twice, citing as reasons not only capital out much debate or delay. Typically, it outflows in the face of interest hikes by was the local branch office that provi- the Bank of England, the Banque de sionally raised the limit, while the France and the German Reichsbank but directorate in Vienna sanctioned the also the need to increase the opportu- decisions only ex post.18 In addition to, nity costs of idle cash and to create in- and independently of, individual incentives for intermediaries to relend in creases, a circular allowed all counter- the markets, as the Nationalbank’s sec- parties to exceed their discount limits retary general underlined: “The ten- subject to the posting of eligible securities sions remain and almost no transactions as additional collateral.19 are completed on the open market. The The ultimate quantitative evidence implementation of exceptional mea- in favor of free lending is provided by sures is justified to combat these condi- the behavior of open market discount tions, which have been ongoing for rates: they increased toward the bank some time now and which need to be rate but never exceeded it. Unlimited considered critical” (OeNB, 1912c). access to central bank refinancing at A final, but crucial aspect of bank the official rate prevented the market policy was the management of moral rate from rising above the official rate.20 hazard. In principle, the origin of the As the Austro-Hungarian Bank accom- crisis – a clearly external event unre- modated the high demand for liquidity, lated to pre-crisis policies of financial its overall lending portfolio ballooned, intermediaries – limited moral hazard mirroring the geographical pattern of to begin with. However, different bank distress (both chart 1). The wide banks’ financial situation very likely definition of eligible collateral for lom- differed in terms of their volume of bard operations already before the crisis liquid reserves: Banks with higher was helpful, as it lowered the probability reserves were more resilient to with- that counterparties would run out of standing a run ex ante. Indiscriminate eligible assets. While in Vienna and free lending during the run could thus Budapest, posted collateral consisted mean endorsing risky liquidity manage- mainly of government bonds (1911: ment strategies ex post and could 87%, 1912: 76% of total collateral), encourage imprudent behavior in the

18 For increases in individual limits, see OeNB (1912b, 1913b). 19 See OeNB (1912a). Of course, banks could have used these securities for advances instead, but haircuts would have been higher. Employing securities as collateral for discounts could thus be advantageous. 20 The “market rate” referred to is the average open market short-term discount rate for prime bills. It was calculated and reported daily in the official stock market report of the Vienna stock exchange.

148 OESTERREICHISCHE NATIONALBANK Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

future. The tool to address moral haz- lend based on the quality of the paper ard was the monitoring scheme the submitted, focusing mainly on the Nationalbank had available through its quality of the other signatures on the credit lists and discount committees. bills, as the signing parties would have While the bank lent generously during had to step in if ever Ústřední banka the crisis, it made clear that help was failed to pay. At the same time, future temporary. Increases in discount limits access to central bank refinancing was were typically granted for three months made conditional on changes in Ústřední only. Whereas the Nationalbank stood banka’s business model. The directors’ ready to prolong the extraordinary strategy seems to have worked, as it credit lines if need be, which it did was able to lower the Ústřední banka’s when some of them came up for re- credit limit by December. In February newal in January 1913, counterparties 1913, the minutes report that the man- were nevertheless under pressure to agement of Ústřední banka had been scale down central bank borrowing as changed and that a program to restruc- quickly as possible. Evidence from the ture the bank was already on its way.22 meetings of directors in Vienna sug- To conclude, the way the Aus- gests that the Nationalbank fine-tuned tro-Hungarian Bank managed the panic its pressure depending on whether it of 1912 appears to be a clean example perceived a counterparty’s position as of Bagehotian lending of last resort. By fundamentally sound or not, relying on lending freely, the Nationalbank pre- information obtained from discounting vented a regional banking crisis from as well as the regular reassessments of escalating into a general market liquid- credit limits. An extreme, yet telling ity crisis. The adherence to the “free case is that of Ústřední banka českých lending” principle finds itself unambig- spořitelen in Prague, which had been uously reflected in the fact that market founded as the central institution for rates never rose above the official rate. Czech savings banks in 1903. The bank Last but not least, the Nationalbank’s had apparently overstretched its risk intimate knowledge of its counterpar- management capacities following a ties allowed it to forestall any risk of large business expansion into Galicia moral hazard by forcing its borrowers and Bukovina. In September 1912, i.e. to adjust their business models if they before the start of the run, the Aus- wanted to preserve their access to tro-Hungarian Bank already turned central bank refinancing in the future. down a request by Ústřední banka for an increase in its credit limit. Rather than 4 Free lending – the exception raising the limit when the crisis hit in rather than the rule October, in November the directors We now turn to an analysis of the even argued that Ústřední banka’s credit Nationalbank’s behavior during other limits should be lowered. But they financial panics. All of these crises in- acknowledged that in the face of the volved some sort of flight to safety that ongoing run, such an approach would had the potential to create a price spiral “lead to a catastrophe,” putting Ústřední and depress the market price of a broad banka’s customers at risk.21 Instead, the range of assets. In this sense, they would Austro-Hungarian Bank continued to have all justified the intervention of a

21 OeNB (1912b, November 19 and December 3). 22 OeNB (1913b, February 4), Compass (1868–1919, volume 1914).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 149 Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

LLR. For some of these crises, quanti- before the outbreak of the crises.23 This tative evidence on restrictions and ra- observation raises the question why the tioning in central bank lending is avail- Nationalbank did not abide by the free able in the form of the spread between lending principle, even though ra- market and official rates as reported for tioning conveys a very negative signal, 1912 in chart 1 above. Chart 2, which undermines confidence in the markets displays market and official rates for the and exerts a potentially severe impact on crises of 1848, 1873, 1924–25 and financial stability as well as the loan sup- 1931, shows that in contrast to 1912, ply to the real economy.24 In the follow- credit rationing is obvious in three of ing section, we argue that external these four episodes – sometimes even constrains, but also self-imposed inter-

Chart 2 Bank and market rates during four ﬁnancial crises 1848 1873 % % March 13: Revolution in Vienna 6 18 8 24 May 22: Convertibility suspended May 13: Suspension of Bank Act

4 12 6 18

4 12 2 6 2 6 0 0 0 0 –2 –6 –2 –6

–4 –12 –4 –12

–6 –18 –6 –18 Jan. Apr. July Oct. Jan. Apr. July Oct. 1923–25 1931 % % % 20 20 15 36 May 11: Creditanstalt Oct. 9: Exchange losses published controls introduced 27 15 15 10 18 5 10 10 9 0 0 5 5 –9 –5 –18 0 0 –10 –27 –5 –5 –15 –36 July 1923 Jan. 1924 July 1924 Jan. 1925 Jan. Apr. July Oct. Oﬃcial discount rate set by the Nationalbank (left-hand scale) Market rate for prime bills Vienna (left-hand scale) Market rate for prime bills Trieste (left-hand scale) Depreciation of currency to par (right-hand scale)

Source: Oﬃcial rates: Lucam (1876), OeNB (1923–1938); market rates: Journal des Oesterreichischen Lloyd (1836–1849), Der Tresor (1872–1918), OeNB (1923–1938), Österreichisches Institut für Konjunkturforschung (1927–1938); depreciation from par: authors’ calculations based on silver prices and exchange rates in Wiener Börse (1786–1858), Wiener Börse (1861–1875) and OeNB (1923–1938). Note: Market rates refer to three-month open market discount rates for prime bills. For 1848, direct evidence is only available for Trieste; rates for Vienna refer to implicit interest rates derived from the price of bills with diﬀerent maturities payable in Vienna as traded in Trieste.

23 Unfortunately, an analogous comparison for 1820 is not possible because market rates are not available for this early period. 24 Friedman and Schwartz (1963) already criticized the Federal Reserve for only faintheartedly supporting banks that suffered panic runs during the Great Depression. For more recent research on the financial stability implications of rationing policies, see Richardson and Troost (2009).

150 OESTERREICHISCHE NATIONALBANK Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

nal constraints partly prevented the ated in areas affected by escalating ten- Nationalbank from lending freely, and sions between the imperial army and we look at some institutional innova- revolutionary groups began to experi- tions used to circumvent these obsta- ence increasing funding difficulties. cles. On the other, the revolts fueled a political confidence crisis that induced a run 4.1 The confidence constraint: on the central bank’s exchange offices, convertibility, statutory limits on as people suspected that Chancellor note issuance and the value- Metternich’s struggling authoritarian storing capacity of central bank regime would ultimately take recourse money to debt monetization to pay for rising As pointed out above, a LLR must be military expenditures. Fearful that paper capable of issuing unlimited amounts of money would lose its purchasing power, cash to prevent bank runs or to calm the public drained specie reserves from incipient runs. Although central banks the Nationalbank’s vaults. cannot become illiquid, the privilege to Following the logic in Martin issue more legal tender does not always (2009), raising official rates might have constitute a sufficient basis for acting as slowed this process, but the National- an effective LLR. Indeed, banknotes – bank’s Governing Board unanimously the cash liabilities issued by the central rejected an increase in the discount rate bank – must themselves be accepted as as largely ineffective during a time a liquid and safe means of payment. If, characterized by civil warfare (Press- on top of a run on banks engendering burger, 1959a). Moreover, the National- outflows of liquid reserves from the bank could not count on the govern- banking system, a run on the central ment, which itself was considerably bank occurs because the public loses its weakened by the ongoing revolts, to confidence in the value of banknotes acquiesce to the imposition of defla- and the central bank is compelled to tionary pressures on the economy, nor convert large quantities of cash into was the government in a position to specie (or foreign exchange), the central help out the Nationalbank with its own bank’s ability to calm a panic by issuing silver reserves. In these critical circum- currency is severely curtailed. Histori- stances, the Nationalbank apparently cally, the only way out of a dual run rationed credit to gain time while was the suspension of note convertibil- desperately trying to find an alternative ity into specie or the introduction of to end the depletion of its reserves. By foreign exchange controls. While halting May 1848, the Nationalbank had run the banking panic, these measures out of options. Following some mar- necessarily triggered sharp deprecia- ginal restrictions on the convertibility tions of the currency. of its notes, it asked the government for The crisis of 1848 constitutes a permission to suspend convertibility on prime example of the confidence con- May 20, 1848. Only the forced exchange straint. The financial repercussions of of paper money, which was declared the revolutionary uprisings which hit two days later, freed the Nationalbank the Austrian Empire in 1848 were from the necessity to impose rationing twofold (Pressburger, 1959a). On the and led to downward pressure on market one hand, banks and merchants situ- rates as the Nationalbank re-expanded

MONETARY POLICY & THE ECONOMY Q3– Q4/16 151 Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

lending.25 At the same time, the value into an IPO. As soon as doubts about of the florin (fl.) dropped dramatically the profitability of the newly founded (chart 2). companies surfaced in April 1873, 1848 was not the only instance stock prices reverted, forcing banks to when the confidence constraint became place margin calls with their deeply binding. There is evidence that the illiquid and partly insolvent borrowers. Nationalbank found it challenging to Borrowers defaulted on their repo loans lend freely during financial crises even from banks while depositors suddenly when convertibility was suspended, began to withdraw their funds from which was the case during the entire banks in May 1873.27 second half of the 19th century until the In an almost immediate reaction to introduction of a “shadow” gold stan- the stock market crash on May 9, 1873, dard in the late 1890s. To increase and the Austrian and Hungarian govern- maintain the public’s confidence in the ments jointly agreed to suspend the value-storing capacity of banknotes, Nationalbank’s reserve requirement on from the 1850s onward, the govern- May 13 in order to decouple the central ment imposed statutory rules governing bank’s credit supply from any formal the cover of banknotes by precious reserve rule. Yet, despite the legal pos- metals. The flipside of these rules was sibility to lend freely, the National- that they limited the Nationalbank’s bank’s policy response to the crisis capacity to expand lending in the case stayed very tentative. The cover ratio of a financial crisis. was effectively overshot in only 18 out The new regime was first tested of the 74 weeks of suspension, which during the Gründerkrach stock market was upheld until October 1874. Fur- crash in 1873. At this time, the National- thermore, the amount by which the bank was operating a system modeled statutory limit of note issuance was ex- on the British Peel’s Act; it limited the ceeded remained negligible through- amount of banknotes to fl. 200 million out. The all-time high of excess supply without silver coverage, while every amounted to fl. 28.5 million and florin issued beyond the fl. 200 million was reached on November 11, 1873 ceiling had to be covered by reserves in (Lucam, 1876), which has to be com- specie.26 The Gründerkrach followed a pared to an aggregate equity base of the pronounced financial boom between banking system of fl. 575 million. 1867 and 1873 that saw a sharp increase Apparently, the Nationalbank refused in joint stock banks in operation. Most to lend on eligible collateral, as the new banks engaged in investment rates paid on prime bills in Vienna banking. To raise the attractiveness of repeatedly exceeded official rates by up their initial public offerings (IPOs), the to 100 basis points even after the sus- banks generously lent against shares pension of cover requirements (chart 2). in the form of repos. These schemes The most striking feature of the allowed investors in the stock exchange period following May 13, 1873, is the to accumulate high leverage, as they seemingly contradictory presence of a needed very little own capital to buy strong skepticism of the Nationalbank

25 After having gradually fallen over the first months of 1848, the circulation of banknotes rapidly increased following the suspension of convertibility. See Jobst and Kernbauer (2016). 26 All relevant legal texts are reproduced in Pressburger (1959a, 1959b). 27 For a more detailed discussion of these dynamics, see Rieder (2016).

152 OESTERREICHISCHE NATIONALBANK Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

and the state vis-à-vis free lending, in writing an account of Nationalbank spite of the suspension of the 1863 Bank policy during the period from 1861 to Act by the government itself. Recur- 1875, its secretary general prided him- ring references and discussions in the self on the fact that the Nationalbank minutes of the Nationalbank’s Governing had behaved as if convertibility had Board meetings suggest that its policy been in place (Lucam, 1876). Unsur- decisions were haunted by the fear that prisingly, the maybe overly strict inter- the cover ratio could soon be fully rein- pretation of this approach led to mass stated. The Governing Board argued failures of around 40% of all Austrian that any too sudden reinstatement joint stock banks. However, in stark could catch the Nationalbank on the contrast to the wild gyrations of the wrong foot, if it were to overextend exchange rate on the various occasions credit during the episode of suspension of economic and political turmoil since and could lead to the loss of the 1848, the florin’s silver value remained Nationalbank’s note-issuing privilege. remarkably stable from 1872 onward At the same time, it is hard to believe throughout the entire crisis period. that this threat needed to be taken seri- When the Nationalbank’s statutes ously, given that the government had were renewed in 1888, the numerical itself taken the initiative to suspend the ceiling on the central bank’s note Bank Act. Rather, the Nationalbank’s issuance was maintained, but was also cautiousness seems to have reflected a amended to allow for temporary trans- general consensus on the importance of gressions of the fl. 200 million limit as rules to guard the stability of the cur- long as the bank kept a proportional rency. The credibility of the National- cover ratio of at least 40% and paid a bank’s commitment to the stability of compensatory tax on the excess issue to the florin had suffered significantly the state (Jobst and Kernbauer, 2016). in 1866 when the government had is- This alteration introduced a rule-based sued significant amounts of government flexibility into the Nationalbank’s paper money to finance the war against policy framework that was known to Prussia.28 As the convertibility of the the public ex ante and that fundamen- florin into silver had been suspended tally alleviated the tradeoffs the since then, adherence to the statutory Nationalbank had faced in 1873. This cover requirements remained the only setting certainly facilitated the swift formal safeguard in place for the value and encompassing policy response to of the Austrian currency. Lifting the the crisis of 1912 described above, as cover requirements thus represented a the Nationalbank made very extensive potentially far-reaching intervention use of the flexibility granted by the that put the public’s confidence in the introduction of the compensatory tax. florin to a test. This also seems to have At end-1912, the sum of taxable been the position of the government, banknotes in circulation was by far the which, when it informed the National- highest since the institution of the rule bank about the temporary suspension in 1888 (Pressburger, 1973). Undoubt- on May 13, 1873, advised it not to abuse edly, the room for maneuver conceded its newly gained freedom.29 When by the statutes of 1888 represents a key

28 See Jobst and Stix as well as Prammer et al. (2016). 29 Memorandum to the Governing Board signed by Finance Minister de Pretis on May 13, 1873, as cited in Pressburger (1959b).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 153 Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

reason for the Nationalbank not to 1920s. While the OeNB tried to post- engage in credit rationing during the pone interest rate increases, it was not turmoil of 1912. in a position to expand its lending too The handling of the crises in 1848, much without attracting heavy criti- 1873 and 1912 highlights the impor- cism from the League’s finance com- tance of the modalities and details mittee and the Bank of England (Kern- woven into the framework of monetary bauer, 1991). From a purely technical policy. Only a very specific combination perspective, the OeNB could have im- of tools and constraints enabled the plemented a “free lending” policy while central bank to walk the delicate tight- maintaining the low level of interest rope between free lending on the one rates and thus could have pushed down hand and maintaining public confi- market rates. However, full allotment dence in the value-storing capacity of at low interest rate levels would have its most important liabilities, Austrian induced a significant increase in the cir- banknotes. Nevertheless, it is worth not- culation of banknotes at a time when ing that the above cases could convey renewed inflation and exchange rate the false impression of a linear evolu- volatility constituted the most pressing tion of the Nationalbank’s LLR role, an concerns of Austria’s international assumption that looking at later crises creditors. cannot substantiate. Following hyper- In another deviation from the linear inflation and the stabilization of the trend, the twin crisis that erupted in Austrian currency in 1922, the new the aftermath of the near-failure of statutes of the re-established Oester Creditanstalt in May 1931 again trig- reichische Nationalbank (OeNB) in 1923 gered public doubts about the OeNB’s lifted the absolute limit on the issuance ability to defend the Austrian schilling’s of banknotes, reduced the proportional fixed exchange rate (Schubert, 1991; cover ratio further and even allowed Stiefel, 1989). Unfortunately, a lack of for transgressions of the proportional comparable market rates makes it im- cover ratio as long as the central bank possible to evaluate whether and to paid a compensatory tax in return what extent mistrust in the value of the (Kernbauer, 1991). Convertibility of Austrian currency curtailed the OeNB’s banknotes was still suspended, and, as free lending capacity following the previously, the OeNB was charged with Creditanstalt crisis in 1931. Yet, the the task of preempting a depreciation of measures taken suggest that the OeNB the currency until the resumption of faced a drastically deteriorating situation. specie payments. Yet, despite the fact The OeNB engaged in unsuccessful that the operational environment of attempts to reduce central bank bor- 1912 had been transformed to include rowing by hiking interest rates, and, even more flexibility on paper, free equally in vain, tried to replenish its lending was again absent during the reserves via loans from the Bank of so-called “franc crisis” of 1924 (chart 2). International Settlements (BIS). Only The OeNB advanced liquidity to the the combination of a complete govern- market, but it could not fully satisfy de- ment guarantee for national and inter- mand without risking a violation of the national creditors’ funds at Creditanstalt, gold standard orthodoxy imposed by on top of an additional state guarantee the League of Nations, under whose on the bills (to be) discounted by the tutelage both the government and the OeNB in conjunction with the intro- OeNB had operated since the early duction of capital controls, proved

154 OESTERREICHISCHE NATIONALBANK Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

enough to halt the run on the Creditan- of the Napoleonic wars. In September stalt. The price to be paid was the loss of 1820, a (temporary) sharp decline in a significant part of foreign exchange the price of government debt put a reserves and the reintroduction of ex- number of highly leveraged Viennese change controls in October 1931. To- bankers into difficulties.30 Part of the gether, the crises of 1924 and 1931 thus leverage had been financed through suggest that additional dimensions be- lombard lending by the Nationalbank, yond the scope of the present study, which now faced a double challenge. such as the prevailing political forces On the one hand, the decline in market and ideas, as well as the (international) prices obliged it to make margin calls, creditor structure, equally shape the which the counterparties could not ful- LLR’s room for maneuver. fill. According to its regulations, nonpayment allowed the Nationalbank to 4.2 Lack of eligible assets and liquidate the collateral, which, how- possible remedies ever, would have precipitated a further Successful free lending requires that decline in the market price, ultimately the counterparties in need of liquidity resulting in a loss to the Nationalbank have sufficient assets at hand that can be when selling (OeNB, 1820, July 14). As sold to the central bank or pledged in a result, the Nationalbank had to tweak collateralized borrowing. In principle, its own rules and suspended margin assets should suffice as long as the calls.31 While this measure allowed it to central bank stands ready to value them keep the level of pre-crisis lending, it at their pre-panic prices. In the 19th did not answer the need of some trou- century setting, bills of exchange did bled houses for additional funding. not pose any complications, as they Now the high haircuts, a useful risk were always valued at 100 minus a management tool ex ante, proved prob- deduction for the discount rate. In the lematic, as borrowing from the Natio case of securities used in lombard lending, nalbank at 75% of market value meant central banks could simply use the pre- that counterparties could not use the panic prices to avoid procyclical tight- remaining 25% as collateral for borrow- ening of credit. ing e.g. from other banks. Ultimately, In practice, however, the situation the constraint was circumvented by in Austria was more complex for several the intermediation of a committee of reasons. First, the Nationalbank was other Viennese bankers as well as a bound by its statutes and regulations. In government guarantee (OeNB, 1820, its collateralized lending operations, September 28). The episode neverthe- the bank’s regulations combined mark- less underlines the constraints imposed to-market pricing (like in the Eurosys- by haircuts and mark-to-market pricing tem today) with significant haircuts. for collateralized lending in a situation The Nationalbank’s first crisis manage- of falling market prices. ment experience in 1820 is particularly The Nationalbank was also subject informative: The 1820 crisis erupted in to stringent regulations on discount lend the context of the long-run recovery of ing, as mentioned in section 2 above. Austrian bond prices in the aftermath For a long time, counterparties in dis-

30 In 1820, the circle of Austrian bankers consisted exclusively of private banking houses. The first joint stock banks emerged only in the 1850s. OeNB (1820, October 5). 31 OeNB (1820, September 21 and October 5).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 155 Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

count operations had to reside in Vienna scribed in section 2, the Nationalbank or a city equipped with a Nationalbank worked with individual credit lines, branch office. In addition, all bills dis- which it increased at its discretion only counted had to be payable in such cities. in cases of acute need. These constraints together excluded a To circumvent these different restric- significant number of agents and bills of tions, the Nationalbank ultimately used exchange from central bank refinanc- several ad hoc tools and options, most ing.32 With the extension of the Nation- of which only partly addressed credit albank’s branch network after the rationing and, moreover, entailed a 1850s, these territorial restrictions transgression of its statutes or some- gradually lost importance. However, times best practice risk management. they appear to have had implications Most interesting in this respect is during the turmoil of 1848. Lacking that the Nationalbank repeatedly re- branches in the provinces, the National sorted to a more systematic alternative bank was initially unable to influence response by setting up support com- local credit conditions in 1848. The mittees (Aushilfs-Comités). These com- constant positive spread of the market mittees were usually formed by repre- rate on prime bills in Trieste over the sentatives of several leading banks in corresponding costs in Vienna (chart 2) Vienna, but also of firms outside the suggests that liquidity in the provinces banking system. They joined forces to was available only in return for the pay- buy bills not admissible at the central ment of a hefty premium that could in bank’s discount window and agreed to fact exceed 150 basis points, even in a lend against collateral outside the financial center like Trieste. central bank’s eligibility circle. The Even if the Nationalbank had sus- Nationalbank in turn contributed to pended its rules and had accepted bills financing the operations of these support payable outside a branch city in 1848, committees either directly by discounting the extension of eligibility would have otherwise ineligible bills once the com- met with a second set of constraints, mittee members had jointly guaranteed namely the interrelated issues of infor- them, or indirectly by agreeing to gen- mation and risk management. Lending erously discount bills presented by the to new agents required knowledge of allied credit institutions. The degree of their quality, i.e. of whether bills would institutionalization of these committees have constituted “good collateral” before varied from time to time, as did the the onset of the crisis. As evidenced by duration of their existence and hence the sophisticated infrastructure for the extent of the Nationalbank’s in- counterparty screening described in volvement. In 1820, for example, the section 2, such information was costly committee constituted only an ad hoc and time-consuming to obtain. Further- association of 12 private banking houses more, the need to manage risk appropri- that mutually guaranteed their liabili- ately also limited the extension of addi- ties to gain access to the discount tional facilities to existing counterpar- window for the purpose of support- ties that could have stood in as inter- ing their ailing peers (OeNB, 1820, mediaries between ineligible counter- September 21). The Vienna Aushilfs- parties and the central bank. As de- Comité of 1873, in contrast, represented

32 See Jobst and Kernbauer (2016). For a similar argument made in the context of the Banque de France, see Bignon and Jobst (2016).

156 OESTERREICHISCHE NATIONALBANK Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

a longer-lasting and more sophisticated unknown. By providing a first system- organization. It published a full-scale atic treatment of banking panics that business program and sourced a perma- hit Austria during the 19th and the first nent equity fund drawn up by Austro- half of the 20th century as well as of the Hungarian firms and private business- respective policies the Austrian central people. The Nationalbank even pledged bank adopted, this contribution has fl.1 million, representing one-eighth of added several observations to the inter- total equity, to cover potential losses national literature on financial crises resulting from lending activities (Com- and has enlarged the pool of experience pass, 1868–1919). to draw from. By construction, the Aushilfs-Comités The analysis focused on episodes served several purposes. First, they where LLR policies addressed the market, minimized the central bank’s counter- more precisely, the entirety of the central party risk because the participants had bank’s eligible discount and lombard given a mutual guarantee; at the same counterparties. Occasionally, however, time, they put the central bank into a the Nationalbank designed, or partici- position to channel credit to borrowers pated in, schemes addressing individual with potentially above-average risk troubled institutions that did not qualify using collateral beyond its statutory for the Nationalbank’s standard crisis requirements. Moreover, by leaving the operations because they did not have a business of rating submitted bills and sufficient amount of eligible assets. collateral to the guaranteeing mem- Typically, such special treatment was bers, the Nationalbank appeared to rationalized by the extraordinary im- have found a clever way to access portance of the respective institution. private information and to shift some of Today, such institutions would be called the additional risk management costs systemically important financial insti- back to the market.33 Finally, the com- tutions (SIFIs). Austrian history features mittees might also have helped several such cases, most notably Boden- coordinate important market players in creditanstalt in 1873 and again in 1929 preventing fire sales and might thereby as well as Biedermannbank, Centralbank have addressed one of the most conse- der Deutschen Sparkassen and Postsparkasse quential collective action problems (Postal Savings Bank) in the mid- associated with banking panics. 1920s.34 Finally, support for Creditanstalt in 1931 also belongs in this category. 5 Conclusions However, the economic (and of course The failure of Creditanstalt in 1931 also political) logic of lifeboats and bail- became notorious beyond Austrian outs is distinct from the LLR opera- borders. Other episodes of banking tions described here and requires very distress in Austria, with the possible detailed evidence on each particular exception of the crisis of 1873, are less case as well as separate treatment. well known or are internationally

33 The definition of “cost-covering fees for the valuation and risk management of less liquid assets that are submitted to the central bank as collateral” is at the center of recent discussions on how to reduce overreliance on central bank lending. See Bindseil (2014). 34 On Bodencreditanstalt, see the ongoing dissertational work of Kilian Rieder. On the interwar bailouts, see Ausch (1968), Stiefel (1989) and Kernbauer (1991).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 157 Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

Four observations about the LLR fidence in the liabilities issued by the operations considered here are in order: Nationalbank itself. However, rather First, while crisis episodes exhibit than announcing defeat, the National- recurring patterns of policy responses, bank produced a number of institu- no clear evolution can be discerned.35 tional innovations designed to circumvent Although the Nationalbank was set up some of these limitations: so-called as a private joint stock company in Aushilfs-Comités, associations of private 1816, several indicators signal that it agents partly funded by the central operated in a broader interest from the bank, played the most prominent role outset. This pursuit of a more general from 1820 onward. They appear to interest is evident e.g. in 1820, when have addressed several concerns, allow- the directors evoked the responsibility ing the central bank to draw on private of their “national institution” to halt the funds and to collect information on crisis, or in 1854, when help for a pro- counterparty and collateral quality; vincial savings bank was justified by they also helped to align incentives and potential contagion effects on other coordinate market agents in order to savings banks.36 A likely explanation for prevent fire sales. As pervasive as they the early inclusion of financial stability were, they have hardly been analyzed in in the Nationalbank’s objective func- the literature. Their rationale, function may be that it never operated as a tioning as well as the extent of their narrowly profit-oriented and competitive success (or failure) would be a worthy corporation but was instead run in the subject for further research, in particular joint interest of a coalition of the most as evidence on similar instruments in important Vienna banking houses and other countries is relatively scarce.37 the government from the beginning. Third, our reading of Austrian bank- Second, our analysis has shown that ing crises underlines the importance of cases of free lending were rare. Con- the microeconomics of last-resort lend- trary to Bagehot’s prescriptions, the ing, hereby echoing the arguments central bank typically rationed credit, made by Flandreau and Ugolini (2014) notably in 1848 and 1873, and probably for the case of the Bank of England. also in 1820 and 1924. As argued above, The Nationalbank scrutinized both the the reason was not that the National- nature of submitted bills and collateral bank did not recognize its responsibility as well as all available information on for the stability of the financial system. individual borrowers. The accumulation Rather, its policies reached the limits of information worked not only to limit imposed by first, its statutes, which the central bank’s risk exposure but forbade some types of operations or set also to reduce moral hazard and to a maximum for the amount of resources provide incentives to limit dependency available for lending; second, a lack of on central bank lending. Equally of information, which would have made interest is the definition of securities lending excessively risky; or third and eligible for lombard lending before and most dramatically, a lack of public con- during a crisis. Here again, our know

35 Unlike argued e.g. in Capie et al. (1994). 36 OeNB (1820, September 21), OeNB (1854). On the instances of support to savings banks in Bratislava and Košice, see Jobst (2014). 37 Some 19th-century central banks used similar arrangements when creating a provincial branch network. See Ugolini (2012) for the case of Belgium.

158 OESTERREICHISCHE NATIONALBANK Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

ledge of historic central bank practices financial crisis during the 40 years is limited, the only regularly cited following the 1873 crash, as the panic evidence being Bank of England’s Gov- of 1912 prominently featured in this ernor John Horsley Palmer’s testimony article is in fact an example of a crisis that, in 1825, the Bank of England had successfully averted. Given central lent “against all sorts of securities with- banks’ desire to avoid financial crises, it out much inquiry as to their nature.”38 is these periods of calm that may contain The evidence on Austrian banking crises the most interesting lessons for today. suggests a different picture; in Austria The controversies that accompanied the definition of eligible collateral was central bank actions worldwide during extensively debated by the Governing recent financial crises inter alia led to a Board of the Nationalbank. Akin to the significant curtailing of the Federal question of individual lending ceilings, Reserve’s possibilities to support indi- the delimitation of eligible collateral is vidual institutions in the 2010 Dodd- a crucial microeconomic aspect of cen- Frank Act, showing that the debate on tral bank lending that has significant the need for and the optimal design of consequences for the effectiveness of LLRs is well and alive. At the core of crisis policy and the extent of moral the current debate are questions of sys- hazard it produces ex post. temic stability, the too-big-to-fail concept, Lastly, one question could not be the cost of bank rescues and the options treated here, namely what feedback for bail-ins as well as the likely negative central bank policies have on the occur- consequences of banks expecting a rence of banking distress. Unlike the government bail-out in the future. U.S.A. with its frequent banking panics Authorities in the 19th as well as early or France and the U.K., which saw 20th century already faced many of these failures or at least the rescues of poten- issues. Their creative solutions offer tially systemically important institu- potentially important insights for today’s tions, Austria did not witness any policymakers.

References Acharya, V., D. Gale and T. Yorulmazer. 2012. Rollover Risk and Market Freezes. In: Journal of Finance 66 (4). 1177–1209. Allen, F., A. Babus and E. Carletti. 2009. Financial crisis: theory and evidence. In: Annual Re- view of Financial Economics 1. 97–116. Allen, F. and D. Gale. 2007. Understanding financial crises. Clarendon lectures in finance. Oxford, UK: Oxford University Press. Almunia, M., A. Bénétrix, B. Eichengreen, K. O’Rourke and G. Rua. 2010. From Great Depression to Great Credit Crisis: Similarities, Differences and Lessons. In: Economic Policy 25 (62). 219–265. Antinolfi, G., F. Carapella, C. Kahn, A. Martin, D. C. Mills and E. Nosal. 2015. Repos, fire sales, and bankruptcy policy. In: Review of Economic Dynamics 18 (1). 21–31. Ausch, K. 1968. Als die Banken fielen – Zur Soziologie der politischen Korruption. Vienna: Europa Verlag. Bagehot, W. 1873. Lombard Street: A Description of the Money Market. London: Henry S. King & Co.

38 Reported as quoted in Wood (1939).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 159 Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

Bignon, V. and C. Jobst. 2016. Economic Crises and the Lender of Last Resort: Evidence from 19 th century France. Mimeo. Bignon, V., M. Flandreau and S. Ugolini. 2012. Bagehot for beginners: the making of lender- of-last-resort operations in the mid-nineteenth century. In: Economic History Review 65 (2). 580–608. Bindseil, U. 2014. Monetary policy operations and the financial system. Oxford: Oxford University Press. Brunnermeier, M., A. Crocket, C. Goodhart, M. Hellwig, A. Persaud and H. Shin. 2009. The Fundamental Principles of Financial Regulation. Geneva Reports on the World Economy 11. Buyst, E. and I. Maes. 2008. Central banking in nineteenth-century Belgium: was the NBB a lender of last resort? In: Financial History Review 15(2). 153–173. Capie, F., C. A. E. Goodhart and N. Schnadt. 1994. The development of central banking. In: Capie, F. et al. (eds.). 1994. The Future of Central Banking: The Tercentenary Symposium of the Bank of England. Cambridge: Cambridge University Press. 1–231. Castiglionesi, F. and W. Wagner. 2012. Turning Bagehot on His Head: Lending at Penalty Rates When Banks Can Become Insolvent. In: Journal of Money, Credit and Banking 44 (1). 201–219. Compass. 1868–1919. Finanzielles Jahrbuch für Österreich-Ungarn. Volume 1877. Vienna: Hölder. Der Tresor. 1872–1918. Vienna. Eichengreen, B. 2015. Hall of mirrors: the Great Depression, the Great Recession, and the uses – and misuses – of history. Oxford: Oxford University Press. Flandreau, M. and S. Ugolini. 2014. The Crisis of 1866. In: Dimsdale, N. and A. Hotson (eds.). 2014. British Financial Crises since 1825. Oxford, UK: Oxford University Press. 76–93. Friedman, M. and A. Schwartz. 1963. A monetary history of the United States, 1867–1960. Princeton, NJ: Princeton University Press. Goodhart, C. A. E. 1999. Myths about the Lender of Last Resort. In: International Finance 2 (3). 339–360. Gorton, G. 2012. Misunderstanding financial crises: why we don’t see them coming. Oxford, UK: Oxford University Press. Gorton, G. 2016. The History and Economics of Safe Assets. In: NBER Working Paper Series 22210. Gorton, G. and A. Metrick. 2012. Securitized banking and the run on repo. In: Journal of Financial Economics 104 (3). 425–451. Gorton, G. and G. Ordoñez. 2014. Collateral Crises. In: American Economic Review 104 (2). 343–378. Jobst, C. 2014. What drives central bank branching? The government, the Austrian National Bank and local lending in the territory of present-day Slovakia 1845–1880. In: Hallon, L., A. Leková and F. Chudják (eds.). Central Banking in Central Europe. Bratislava. 14–27. Jobst, C. and H. Kernbauer. 2016. The Quest for Stable Money. Central Banking in Austria, 1816–2016. Campus. Frankfurt. New York. Jobst, C. and H. Stix. 2016. Florin, crown, schilling and euro: an overview of 200 years of cash in Austria. In this volume. Jobst, C. and S. Ugolini. 2016. The coevolution of money markets and monetary policy, 1815– 2008. In: Bordo, M. et al. Central banks at a crossroads. Cambridge: Cambridge University Press. 145–194. Journal des Oesterreichischen Lloyd. 1836–1849. Trieste.

160 OESTERREICHISCHE NATIONALBANK Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

Kernbauer, H. 1991. Währungspolitik in der Zwischenkriegszeit – Geschichte der öster reichischen Nationalbank von 1923 bis 1938. Vienna: Oesterreichische Nationalbank. Kindleberger, C. and R. Aliber. 2011. Manias, panics and crashes: a history of financial crises. Basingstoke: Palgrave Macmillan. k. k. Statistische Zentralkommission. 1915. Österreichische Statistik N.F. 12 (2). Statistik der Sparkassen für das Jahr 1912. Vienna: Gerold. Kövér, G. 2015. Information transfer in the local discount committees of the Austro-Hungarian Bank 1878–1914. Paper presented at the Workshop The central bank balance sheet in the (very) long run. October 1. Vienna: Oesterreichische Nationalbank. Lucam, W. 1876. Die Oesterreichische Nationalbank während der Dauer des dritten Privilegiums. Vienna: Manz. Martin, A. 2006. Liquidity Provision vs. Deposit Insurance: Preventing Bank Panics without Moral Hazard. In: Economic Theory 28 (1). 197–211. Martin, A. 2009. Reconciling Bagehot and the Fed's Response to September 11. In: Journal of Money, Credit and Banking 41 (2–3). 397–415. Martin, A., D. Skeie and E.-L. Thadden. 2014. Repo Runs. In: Review of Financial Studies 27 (4). 957–989. Michel, B. 1976. Banques & banquiers en Autriche au début du 20e siècle. Paris: Presses de la Fondation Nationale des Sciences Politiques. Naqvi, H. 2015. Banking crises and the lender of last resort: How crucial is the role of information? In: Journal of Banking & Finance 54 (5). 20–29. Neue Freie Presse. 1864–1939. Vienna. Oehmke, M. 2014. Liquidating illiquid collateral. In: Journal of Economic Theory 149. 183–210. OeNB. 1820. Protokolle der Direktoriumssitzungen (Minutes of the meetings of the Governing Board). July 14. September 21. September 28. October 5. OeNB Bank History Archives. OeNB. 1854. Protokolle der Direktoriumssitzungen (Minutes of the meetings of the Governing Board. September 14. OeNB Bank History Archives. OeNB. 1912a. Circular. December 3. OeNB. 1912b. Minutes of the meetings of the directorate in Vienna (Protokolle der Direktion in Wien). September 24. October 29. November 19. December 3. OeNB Bank History Archives. OeNB. 1912c. Protokolle der Generalratssitzungen (Minutes of the meetings of the General Council). November 28. OeNB Bank History Archives. OeNB. 1913a. Annual general meeting of the Austro-Hungarian Bank (Jahressitzung der General versammlung der Oesterreichisch-ungarischen Bank). OeNB Bank History Archives. OeNB. 1913b. Minutes of the meetings of the directorate in Vienna (Protokolle der Direktion in Wien). January 14. February 4. OeNB Bank History Archives. OeNB. 1923–1938. Governing Board circulars (Mitteilungen des Direktoriums). OeNB Bank History Archives. Österreichisches Institut für Konjunkturforschung. 1927–1938. Monatsberichte des Öster reichischen Institutes für Konjunkturforschung. Vienna. Prammer, D., L. Reiss and W. Köhler-Töglhofer. 2016. The financial relations between the Nationalbank and the government. In this volume. Pressburger, S. 1959a. Das österreichische Noteninstitut 1816–1866: Die Vorgeschichte, 1816–1862. Vienna: Oesterreichischen Nationalbank. Pressburger, S. 1959b. Das österreichische Noteninstitut 1816–1866: 1868–1877. Vienna: Oester reichischen Nationalbank. Pressburger, S. 1973. Das österreichische Noteninstitut 1816–1866: 1901–1913. Vienna: Oester reichische Nationalbank.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 161 Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931

Richardson, G. and W. Troost. 2009. Monetary Intervention Mitigated Banking Panics during the Great Depression: Quasi-Experimental Evidence from a Federal Reserve District Border, 1929–1933. In: Journal of Political Economy 117 (6). 1031–1073. Rieder, K. 2016. A Historic(al) Run on Repo? Causes of Bank Distress During the Austro- Hungarian Gründerkrach of 1873. In: Working Paper in Economic History 2016. 1–79. Schubert, A. 1991. The Credit-Anstalt crisis of 1931. Cambridge: Cambridge University Press. Stiefel, D. 1989. Finanzdiplomatie und Weltwirtschaftskrise: die Krise der Credit-Anstalt für Handel und Gewerbe 1931. Frankfurt: Knapp. Thornton, H. 1802. An Enquiry into the Nature and Effects of the Paper Credit of Great Britain. London: Hatchard. Ugolini, S. 2012. L’Invention d’un Système Monétaire National: Banques d’Émission, Supervision Bancaire et Développement Financier en Belgique 1822–1872. Available at SSRN : http://dx.doi.org/10.2139/ssrn.2033671 (retrieved on August 12, 2016). Wiener Börse. 1786–1858. Wechsel-Cours in Wien. Vienna. Wiener Börse. 1861–1875. Cours-Blatt des Gremiums der k. k. Börse-Sensale. Vienna. Wood, E. 1939. English theories of central banking control, 1819–1858: with some account of contemporary procedure. Cambridge, MA: Harvard University Press.

162 OESTERREICHISCHE NATIONALBANK The changing role of macroprudential policy in Austria after World War II

The need for and implementation of macroprudential supervision was a key lesson from the Sophia Döme, global financial crisis of 2008. However, historical protocols, legislation, policy agreements as Stefan W. Schmitz, well as the literature bear witness of a longer history of macroprudential policy in Austria: it Katharina Steiner, predates the crisis by about 60 years during which it gradually evolved. We argue that careful Eva Ubl1 analysis of this history provides interesting insights for current policymakers. We identify the following key lessons from analyzing the changing legal nature, motivation and effectiveness of macroprudential policy in Austria: first, macroprudential policy requires a sound legal basis; second, measures have to be quite intrusive to effectively curtail the build-up of systemic risk. Less intrusive measures become effective above all by increasing the shock-absorbing capacity of the financial system, once sytemic risks materialize. JEL classification: E50, E60, G20, N24 Keywords: history of Austrian macroprudential policy, credit control agreements, minimum liquidity requirements, Limes, Kreditplafond, capital buffers

Macroprudential supervision was for- We argue that although the term “mac- mally introduced in Austria in 2014 as a roprudential” itself denotes a rather re- key lesson from the financial crisis of cent concept, macroprudential policy as 2008. Macroprudential supervision is a such does have a history from which we mainly national responsibility within can learn. This paper summarizes the an increasingly harmonized and cen- history of the changing role of macro- tralized system of financial supervision prudential policy in Austria from the in the European Union (EU) and, even end of World War II (WW II) in 1945 more so, the euro area. The Oesterre- to 2016 and draws lessons for the future. ichische Nationalbank (OeNB) is en- This paper is structured as follows: trusted with a major responsibility for Section 1 defines macroprudential pol- macroprudential supervision in Austria. icy for the purpose of this paper. In Expectations regarding the impact of section 2, we discuss the conceptual macroprudential policy on reducing the framework for macroprudential policy frequency and costs of financial crises in Austria. Section 3 outlines the evo are high, but skeptics have warned that lution of the relevant legal framework. the high hopes in the new tools of mac- Section 4 and section 5 provide a chronol- roprudential supervision might be over- ogy of macroprudential measures taken stated as little is known about their ef- between 1945 and 2016. Finally, sec- fectiveness and transmission channels tion 6 concludes. (Dudley, 2015).

1 Oesterreichische Nationalbank, Financial Stability and Macroprudential Supervision Division, zsofia.doeme@ oenb.at, [email protected], [email protected], [email protected]. The views expressed in this paper are exclusively those of the authors and do not necessarily reflect those of the OeNB or the Eurosystem. The authors would like to thank the referee, Karl Socher, Walpurga Köhler-Töglhofer and the participants in a workshop among the authors of this volume for their very useful comments as well as Irene Popenberger, Susanne Stein- acher and Rita Schwarz for language advice and editing support. Thanks also go to the OeNB’s library and Bank Refereed by: History Archives for their invaluable support throughout this research project. All remaining errors are those of the Eric Monnet, authors. Banque de France

MONETARY POLICY & THE ECONOMY Q3– Q4/16 163 The changing role of macroprudential policy in Austria after World War II

1 Definition of macroprudential bility – financial stability considerations policy played a secondary role. These objec- The term macroprudential – as it is used tives were addressed predominantly by today2– was coined by the Bank for credit policy, and this approach gave International Settlement’s Cooke Com- rise to macroprudential policy long mittee, which mentioned the term for before its formal introduction. This the first time in the minutes of its meet- happened not only in Austria, but in ing on June 28 and 29, 1979 (Clement, several developed economies. 2010). In the EU, a legal definition of In the following sections, we ana- macroprudential supervision was intro- lyze the most relevant instruments that duced only in 2013 with the implemen- were used to meet the macroprudential tation of the Capital Requirements policy objectives of mitigating and pre- Directive IV (CRD IV) in response to venting excessive credit growth, lever- the global financial crisis of 2008. Mac- age and maturity mismatches. We roprudential supervision was set up to study credit control agreements (in- supplement microprudential supervi- cluding credit ceilings) in section 4.1, sion and monetary policy by promoting the 1970s Limes, a bank lending limit, financial stability and reducing systemic in section 4.2, and the Gentlemen’s and procyclical risks to the financial sys- Agreement between banks and the tem and the real economy.3 The Euro- OeNB regarding the sterilization of pean Systemic Risk Board (ESRB, 2013) short-term capital inflows in section specified five intermediate objectives to 4.3, the predecessors of modern mac- operationalize its ultimate objective of roprudential measures (e.g. measures preserving financial stability: (1) miti- that curbed foreign currency loans) in gating and preventing excessive credit section 5.1, the sustainability package growth and leverage, (2) mitigating and in section 5.2, and finally the instru- preventing excessive maturity mis- ments of modern macroprudential super- matches, (3) limiting direct and indirect vision in section 5.3. exposure concentration within the financial system, (4) limiting the systemic 2 The conceptual framework of impact of misaligned incentives, and (5) macroprudential policy after strengthening the resilience of the finan- WW II cial infrastructure. Up to the 1980s, macroprudential pol- We use these intermediate objec- icy in Austria consisted of credit con- tives to retrospectively classify macro- trol agreements (comprising qualitative prudential policy measures taken after and quantitative credit controls), the 1945. Objectives similar to the above- Gentlemen’s Agreement between Aus- mentioned intermediary objectives (1) trian banks and the OeNB, and a limit and (2), in particular, played an im- to bank lending called the Limes. portant role in Austrian economic The first objective associated with policy after WW II. The underlying this toolset was to complement monetary motivation at the time was to achieve policy. If policymakers expected that the overarching objective of price sta- raising interest rates and/or minimum

2 Macroprudential supervision refers to the supervision of the financial system as a whole. See Eidenberger et al. (2014, p. 84.): “The objective of macroprudential supervision is to contribute to the stability of the financial system as a whole, which requires strengthening the resilience of financial intermediaries and of the financial infrastructure, and limiting the build-up of systemic risks in the economy (e.g. house price bubbles).” 3 Article 13 (1) of the Federal Act on the Institution and Organisation of the Financial Market Authority.

164 OESTERREICHISCHE NATIONALBANK The changing role of macroprudential policy in Austria after World War II

reserve requirements would be insuffi- driven by capital accumulation and pro- cient to fight inflation, excessive credit ductivity growth (Solow, 1956; Swan, expansion, the misallocation of credit in 1956). Austria had a very low capital the economy and/or excessive maturity stock after WW II (Seidel, 2005). In- transformation (Klauhs, 1971) or if they come and savings were low, too. As a feared these instruments would weigh consequence, capital – and often capital too heavily on bank profitability (Minis- goods, too – had to be imported. Qual- try of Finance, 1969, p. 23), they turned itative credit controls aimed at shifting to macroprudential policy to provide the allocation of scarce capital toward additional instruments that would allow productive, growth-enhancing invest- them to curb the growth of credit, ment at the expense of consumption monetary aggregates and inflation. and speculative investment (including The second objective was to prevent speculative inventories). To address the short-term disturbances to the financial ensuing current account deficits, quali- system caused by volatile short-term capital tative credit controls also favored tour- flows; these could induce an excessive ism and export credit. The controls increase or decrease of money and were justified by the belief that the credit supply and/or excessive cur- state was better able to allocate funds rency and maturity mismatches in the toward productive investment than the economy, which would endanger price financial markets. Monetary policy stability or economic growth (Schmitz, complemented qualitative credit con- 1969).4 For example, the central bank trols through preferential liquidity pro- disincentivized banks from attracting vision for investment and export loans. new foreign currency deposits by non- The objectives of macroprudential residents by imposing high minimum policy measures during that period reserve requirements on such deposits were closely linked to those of mone- rather than by reducing interest rates. tary policy (Kelber and Monnet, 2014). Macroprudential policies in the form Banks were considered to be able to of credit control agreements and lend- create money through the extension of ing ceilings were common between credit or shifts from time to sight 1945 and the 1980s across the devel- deposits. As a consequence, the terms oped economies (Goodhart, 1989). The “money supply” and “credit growth” policy toolset often included quantita- were used interchangeably, and it was tive credit ceilings and the qualitative not considered necessary to distinguish steering of credit expansion toward monetary policy objectives from sys- exporting and manufacturing sectors temic risk considerations. at the expense of consumer credit. Regarding the transmission channel, Schmitz (1969) provides an overview of the conceptual framework applied by the measures implemented to curb economists and policymakers rested on short-term capital flows (similar to the a quantitative perspective of bank 1971 Gentlemen’s Agreement, see be- balance sheet management (Goodhart, low) in various developed countries. 1989). An increase of the share of The conceptual framework of quali- highly liquid assets in total assets would tative credit controls rested on the imply a reduction of the share of less neoclassical theory of growth that was liquid assets (loans). In aggregate per-

4 In the period from 1958 to 1968, increases in international reserves accounted for 80% of the increase of central bank money in Austria (Schmitz, 1969).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 165 The changing role of macroprudential policy in Austria after World War II

spective, considerations were based on that are added on top of micropruden- the money multiplier. tial capital requirements. Bankers, in contrast, based their The transmission channel of modern conceptual framework on price-based macroprudential policy instruments is dynamic balance sheet optimization based (1) ex ante on price-based dy- (Klauhs, 1964, p. 40): ceteris paribus, namic balance sheet optimization and liquidity and capital requirements that (2) ex post on containing the sys- constitute economically binding con- tem-wide effects of shocks. Additional straints would increase the input costs of layers of binding macroprudential mea- bank intermediation and reduced bank sures create opportunity costs by forcing profitability. Bank management would banks to deviate from their individually react by adapting the structure of assets optimal asset and liability pricing and and liabilities as well as their pricing.5 balance sheet structure. This, in turn, After the financial crisis of 2008, should internalize negative externalities the close link between macroprudential and avoid systemic risk. If risks material- policy and monetary policy was aban- ize nonetheless, the additional buffers doned and systemic risk became the new aim at containing the repercussions focus.6 The new macroprudential policy within the financial system by increasing framework builds on the insight that the the shock absorption capacity of individ- soundness of individual financial insti- ual institutions. tutions – i.e. their fulfillment of microprudential capital requirements – is not 3 The legal framework for macro- sufficient to ensure systemic stability prudential policy after WW II (Eidenberger et al., 2014). The concep- The first legal framework for macropru- tual framework distinguishes the dential policy, the Kreditwesengesetz cross-sectional and time dimensions of (KWG, Austrian credit services act), systemic risks. Examples of the former was introduced in Austria in 1979, include direct and indirect contagion almost 30 years after the first credit due to the interconnectedness of finan- control agreement had entered into cial institutions, and incentive problems force in Austria in 1951. The Reichs- of financial institutions (e.g. problems kreditwesengesetz (RKWG, Reich Bank- caused by emergency liquidity assis- ing Act), which had been introduced in tance, implicit government guarantees Austria by regulation after the An- and the tax subsidy of debt). Along the schluss (RKWG, 1938), remained in time dimension, excessive credit growth force in Austria after the end of WW II and the procyclicality of credit growth (as amended by RKWG, 1939). Consti- constitute the dominant forms of sys- tutional concerns rendered the RKWG temic risk. The main instruments used 1939 all but inapplicable, however.7 In to address these risks are capital buffers fact, Austria had no capital and liquid-

5 Bankers also feared that the relative increase in the costs of bank loans would drive up the market shares of nonbank funding sources, such as insurance companies, instalment payments and promissory notes. Later macroprudential measures also encompassed the first two alternatives. Market-based funding remained very low in Austria until the early 2000s. 6 The literature on systemic risk has grown substantially since the financial crisis; see Galati and Moessner (2011) for a literature survey. 7 The explanatory notes on the first draft of the Austrian KWG (Ministry of Finance, 1955) offer two explanations for these constitutional concerns: first, the regulatory powers of the supervisory authority (i.e. the Ministry of Finance in consultation with the OeNB) to define specific capital and liquidity ratios were regarded excessive under Austrian constitutional and administrative law; second, the way the RKWG 1938 had been introduced in 1938 contributed to constitutional concerns.

166 OESTERREICHISCHE NATIONALBANK The changing role of macroprudential policy in Austria after World War II

ity regulation until 1979; the voluntary the legal foundations for macropruden- credit control agreements between tial policy. It empowered the Ministry Austrian banks and the Ministry of of Finance – in consultation with the Finance in consultation with the OeNB OeNB – to temporarily impose limits aimed at filling this void from 1951 on- on the growth of credit extended by ward (Heller, 1980). In addition to domestic credit institutions to domes- constitutional concerns the poor state tic nonbanks, if credit growth was of the Austrian banking sector further deemed excessive and, if restrictive hampered the execution of the RKWG. monetary policy had proved insufficient After 1945, the share of nonperform- in addressing the problem of excessive ing assets in total banking assets was so credit growth. Article 22 specified high and bank capitalization was so low quantitative limits for ministerial dis- (effectively negative in aggregate) that cretion in terms of ratios of loan expan- banks were not even required to pub- sion to increases in liabilities or capital lish balance sheets until 1955.8 Two at- (liability-side credit control). However, tempts, in 1955 and 1969, to regulate first the Ministry of Finance had to Austrian banks failed (Ministry of seek a voluntary agreement with the Finance, 1955; 1969). The latter at- various associations of Austrian credit tempt would also have included legal institutions and the OeNB. If these vol- foundations for macroprudential policy. untary agreements were not reached The KWG 1979 finally provided a within one month, the Ministry of sound legal basis for macroprudential Finance, in consultation with the OeNB, policy in Austria.9 The explanatory was entitled to issue a directive on the notes on the KWG 1979 explicitly matter in question. Similarly, if consent mention the objective of establishing on an extension to existing credit con- legal foundations for credit control trol agreements was not reached within agreements to maintain economic sta- four weeks, the Ministry of Finance bility (Nationalrat, 1979). Article 12 could issue a corresponding directive. KWG 1979 specified minimum capital Such directives could remain in force requirements (4% of total liabilities mi- for a maximum of 16 months, and nus covered bonds and very highly liq- banks were fined for noncompliance uid assets). Article 11 provided the legal with the provisions. In addition, Arti- foundations for minimum liquidity re- cle 22 (10) included provisions for as- quirements, which had to be defined by set-side credit controls (e.g. explicit the Minister of Finance within certain limits on the growth rates of certain as- qualitative and quantitative limits (up sets). These were subsidiary to liability- to 35% of total assets) specified in Arti- side measures, however – the explana- cle 11. Curbing excessive credit growth tory notes argued that asset-side constituted one of the motivations for measures have an immediate effect this provision. Article 22 finally laid on aggregate demand and the money

8 The Reconstruction Act 1955 required banks to publish their balance sheets for the first time since 1945 as of year-end 1954 and an aggregate profit and loss statement for the entire decade from 1945 to 1954. Banks had to reach a capital-to-liabilities ratio of 10% by 1964. However, banks whose capitalization was below 4% of liabilities were given additional time until 1980 in the 1969 draft of the KWG (Ministry of Finance, 1969). 9 The OeNB would have preferred the legal framework for macroprudential policy to be included in an amendment to the Nationalbank Act 1955; the OeNB’s subcommittee on open market operations and minimum reserve policy suggested a draft of a new Article 43a NBG 1955 that assigned the power to issue directives to the OeNB in consultation with the Ministry of Finance (OeNB, 1968).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 167 The changing role of macroprudential policy in Austria after World War II

supply (rather than a lagged effect like IMF, 2011) and in academia (Galati and liability-side measures) (Nationalrat, Moessner, 2011, for an overview). In 1979, p. 49). the EU, the Capital Requirements It is a historical irony that the last Directive IV (CRD IV, Directive credit control agreement expired in 2013/36/EU) and the Capital Require- June 1981, only two years after a sound ments Regulation (CRR, Regulation legal foundation for these instruments No 575/2013) provided a new legal had finally been established. The expiry framework for macroprudential policy. came about because no agreement on The former was transposed into Aus- the extension could be reached be- trian law with the 2013 amendments tween the Ministry of Finance, the to the Austrian Banking Act (section 5 OeNB and the Austrian credit institu- on macroprudential supervision), the tions. The Ministry of Finance had Financial Market Authority Act, and never made use of its new macropru- the Nationalbank Act.11 The reform dential powers under the KWG 1979. assigned the Austrian Financial Market At the end of the three decades it had Authority (FMA) as macroprudential taken to establish legal foundations for authority, established the Financial macroprudential policy in Austria, Market Stability Board (FMSB) as pol- macroprudential policy measures fell icy forum and delegated the tasks of out of favor around the world with both prospective systemic risk analysis, pol- central bankers and policymakers icy development and FMSB administra- (Goodhart, 1989; Elliot et al. 2013). As tion to the OeNB.12 a consequence, Article 22 of the KWG Beyond the Austrian Banking Act 1979 was deleted in the 1986 amend- and its forerunner versions, the National- ment to the KWG, only seven years af- bank Act and the Foreign Exchange Act ter it had been introduced. Lawmakers helped shape the legal framework of did not even consider it necessary to macroprudential policy in Austria after provide arguments for its deletion in WW II. The Nationalbank Act 1955 the explanatory notes on the amend- provided the legal foundations for mon- ment (Nationalrat, 1986, p. 36).10 etary policy in Austria and was addi- It was only 28 years later, in 2014, tionally used for targeting intermediary that Austria introduced a comprehen- objectives of macroprudential policy. sive legal and institutional framework First, in Article 43, it empowered the for macroprudential policy. In the af- OeNB to require banks to hold mini- termath of the financial crisis of 2008, mum reserves, which in composition macroprudential policy gained renewed resembled the liquidity requirements of attention in international policy fora the previous credit control agreements. (Financial Stability Board, BIS and Article 43 was inter alia intended as a

10 The KWG 1986 granted the Ministry of Finance the power to increase minimum capital requirements for all banks if this were in the economic interest of a well-functioning banking system (Article 12 (2) KWG 1979, as amended by Federal Law Gazette No. 325/1986). In preparation for Austria’s accession to the European Economic Area (1994), the new Austrian Banking Act 1993 (Bankwesengesetz – BWG) transposed the European Community banking directives into Austrian law. The BWG 1993 restricted the Ministry of Finance’s power to raise minimum capital requirements for all banks by setting an upper limit of 0.5 percentage points. This power could be interpreted as a macroprudential instrument, but it was never used. 11 See Federal Law Gazette No. 184/2013. 12 For details concerning the legal and institutional setup of macroprudential supervision in Austria, see Eidenberger et al. (2014).

168 OESTERREICHISCHE NATIONALBANK The changing role of macroprudential policy in Austria after World War II

legal foundation for macroprudential OeNB had to be notified of foreign cur- policy (Nationalrat, 1955). In case vol- rency imports and exports as well as of untary agreements could not be the disposal of foreign currencies above reached, the OeNB had the legal power a specific value, and all foreign cur- to impose similar measures. In addi- rency had to be offered for exchange to tion, the OeNB had strong arguments the OeNB. All foreign currency trans- to nudge banks toward a voluntary actions were subject to approval by the agreement; it could exclude credit in- OeNB. After 1953, the OeNB com- stitutions without giving reasons from menced a gradual and stepwise liberal- discount transactions and lombard ization process, making the approvals loans. This discretion also allowed the less restrictive. In the first phase OeNB to sanction credit institutions (1954–59), it liberalized current ac- for breaches of the Limes agreements. count transactions with countries of The 1969 amendment to the National- the Organisation for European Eco- bank Act introduced specific rules on nomic Co-operation (OEEC), Canada minimum reserve requirements for for- and the U.S.A. In the second phase eign currency deposits by nonresidents (1959–63), the OeNB granted a gen- (Federal Law Gazette No. 276/1969). eral approval for capital account trans- The relatively high ratios of up to 25% actions to nonresidents and eased the of the stock volume and up to 50% of approval practice for Austrians. The the increase in such liabilities was moti- OeNB implemented the liberalization vated by the then-applicable Gentle- process by issuing official announce- men’s Agreement that aimed to contain ments according to the Foreign Ex- the negative effects of high short-term change Act 1946 without seeking capital inflows via the banking sector. changes to the Act itself, which ensured During the 1980s, market- and price- a high degree of flexibility for the based monetary policy instruments re- OeNB in tightening capital controls if placed arm-length or quantitative in- and when necessary. This was the case struments (including the selective allo- in May 1971, when the OeNB reintro- cation of liquidity by the OeNB to duced the requirement of individual commercial banks). When the Nation- OeNB approval for bank deposits by albank Act was adapted to the require- nonresidents, and in November 1972, ments of the Economic and Monetary when it did so for other forms of capital Union (EMU) in 1998, the amend- imports (requirement in place until ments reflected this development. The 1976). However, vis-à-vis banks, the OeNB’s monetary policy instruments OeNB largely refrained from imposing at that time consisted of the toolset quantitative limits on capital imports or which had been developed by the Euro- exports and from tightening its ap- pean Monetary Institute (EMI). It in- proval practice significantly. Instead, it terpreted minimum reserve require- aimed at driving a wedge between the ments solely as instruments to ensure a interest paid on deposits in Austria and stable demand for central bank money that paid on deposits abroad through and to smooth money market interest the Gentlemen’s Agreement. Similarly, rates (ECB, 2011). quantitative restrictions on capital im- In addition, the Foreign Exchange Act ports via purchases of domestic bonds 1946 (Devisengesetz) imposed strict and equity by nonresidents from Austrian capital controls in Austria starting from residents were implemented in the 1946 (Mooslechner et al., 2007): the form of amendments to the (voluntary)

MONETARY POLICY & THE ECONOMY Q3– Q4/16 169 The changing role of macroprudential policy in Austria after World War II

credit control agreements (October 4.1 Credit control agreements 1972). Moreover, the OeNB required Credit control agreements addressed individual OeNB approval for pur- systemic risks in terms of excessive total chases of real estate and for loans pro- credit growth14 and misallocation of credit vided by nonresidents (November in the economy. They were in force be- 1972). The voluntary measures proved tween 1951 and 1982 and consisted of effective, and the number of applica- three components. Component (1) tions for approval of capital imports re- included qualitative credit controls, mained low. The third and final liberal- which e.g. allowed banks to lend only ization phase commenced in 1981 and for sustainable purposes or imposed ended in 1991, when capital controls restrictions on the advertising of con- were finally abolished in Austria. In sumer loans. In contrast, components principle, capital controls can still be (2) and (3) were often referred to as used as macroprudential tools by coun- quantitative credit controls. They com- tries outside the EU (Habermeier et al., prised liability-side controls either in 2011). However, we do not include the form of minimum liquidity require- them in our account of macropruden- ments (component (2)) and/or a credit tial policy measures in Austria after ceiling (Kreditplafond, component (3)). 1945, because the Gentlemen’s Agree- Additionally, as of 1973, the OeNB ment was the main instrument used to pushed through more intrusive as- avoid short-term capital inflows via set-side controls (i.e. the Limes), which banks. The tightening of capital con- were formally not part of the credit trols merely imposed notification re- control agreements (see section 4.2). quirements that enabled the OeNB to These policies targeted the banking monitor compliance with the Gentle- sector, above all, but their scope and men’s Agreement and prevented the the quantitative requirements were circumvention of macroprudential pol- adapted when necessary to also include icy measures via capital imports by insurance companies and instalment nonbanks. companies. Qualitative credit controls targeted 4 Macroprudential measures the sustainability of credit supply by from 1945 to 1982 avoiding a misallocation of credit in the This section provides a chronology of economy. They stipulated that banks macroprudential policy milestones from were only allowed to grant loans for 1945 to 1982 in Austria.13 It describes economically justified purposes, i.e. the economic nature of each measure, productivity-enhancing investments or its motivation (e.g. to avoid excessive exports, and not for speculative use credit growth or a currency mismatch) (excessive inventories or foreign cur- and an ex post assessment of its effec- rency speculation) or unsustainable tiveness. consumption (OeNB, 1951, April 6).

13 This section does not deal with the activities of the commission for the allocation of credit in the economy (Kom- mission zur Lenkung des öffentlichen und privaten Kredits), which was established in July 1945 (Kreditlenkungs- gesetz 1945), as the authors consider this commission the instrument of a postwar command economy rather than a macroprudential policy instrument. 14 The scope of the definition of credit narrowed over time; initially, all loans were included except short-term interbank loans. In 1957, the reformulation also excluded all interbank loans to banks covered by the agreement, export loans and some other smaller categories.

170 OESTERREICHISCHE NATIONALBANK The changing role of macroprudential policy in Austria after World War II

The first type of credit policy imple- liquidity requirements (April 1951), but mented in Austria along these lines later they also included provisions re- comprised an agreement the Ministry garding a credit ceiling (October 1951) of Finance, in consultation with the and capital-based limits on credit OeNB, reached with two banking sec- growth (1957). tors, namely joint stock banks (Aktien- The minimum liquidity requirements banken) and smaller private banks were defined as a specific ratio of liquid (Verband der Banken and Bankiers). It assets to bank liabilities (liability-side took until 1960 for credit control credit controls) and thus also had an agreements to cover all Austrian bank- impact on maturity transformation in ing sectors.15 the banking system. The affected banks An ex post assessment of qualitative had to hold 30% of their liabilities in credit controls is difficult, but it is un- certain classes of liquid assets as of Jan- likely that they were particularly effec- uary 1952: 10% of first-grade liquidity, tive. The OeNB did not gather data e.g. cash or deposits with the OeNB; regarding banks’ compliance with these 20% of second-grade liquidity, e.g. controls. Early critics pointed out that checks, rediscountable bills of exchange it was almost impossible to assess to or discountable government bonds what extent loans were economically (OeNB, 1951, April 6). Supporting justified, and even banks were said to government financing was no explicit be overburdened with the qualitative objective of the minimum liquidity re- criteria regarding credit controls (e.g. quirements, because discountable gov- Tichy, 1965). However, over time a ernment bonds were only covered by consensus emerged that loans for pro- the second-grade liquidity require- ductivity-enhancing investments and for ments and the requirements covered promoting exports were sustainable, many other asset classes. Banks circum- while consumer loans and loans for vented the credit control agreements speculative purposes were not. In 1975, by substituting loans by bills of ex- qualitative controls for consumer loans change, which they rediscounted at the were partly repealed to spur consump- OeNB to increase their first-grade tion and reintroduced in 1977 in the liquidity. This, in turn, provided addi- light of sharply increasing import-based tional scope for credit expansion with- consumption. That indicates that poli- out breaching the agreement. Liquidity cymakers expected some effectiveness distribution across decentralized sec- in terms of allocating credit. However, tors and banking groups constituted as the effect on excessive credit growth another implementation challenge. The remained negligible, the OeNB com- centralization of liquidity incentivized plemented the qualitative controls with the apex or parent institutions to inten- quantitative credit controls only eight sify maturity transformation and expand months after the first credit control their loan portfolios. They could profit agreement had entered into force. from double liquidity leverage (inter- Quantitative credit controls aimed bank loans counted toward liquid as- at mitigating excessive credit growth, sets) and gained a competitive advan- leverage and maturity mismatches. Ini- tage over less complex banking organi- tially they consisted only of minimum zations (Klauhs, 1964).

15 The arrangement with the regional mortgage banks and the Raiffeisen credit cooperatives was only implemented in 1960.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 171 The changing role of macroprudential policy in Austria after World War II

If minimum liquidity requirements sions of the credit control agreements were binding, i.e. had prevented banks (minimum liquidity requirement and from achieving their individually opti- credit ceilings). Banks which did not mal pricing and balance sheet struc- comply with the minimum liquidity re- ture, they would have caused opportu- quirements had to reduce lending until nity costs. Banks could either react by compliance was restored.17 In addition, accepting declining profit margins or they had to pay a fine of 2% of the charging higher interest rates for loans. liquidity gap to the Treasury.18 In the This mechanism should help curb beginning of the 1970s, the rules for credit expansion (Klauhs, 1964). fines were loosened for the first time. In October 1951, the OeNB pushed In case of noncompliance with the for an extension of the credit control credit ceiling caused by a lowering of agreements by specifying a credit ceiling the ceiling, fines were suspended. The (OeNB, 1951, October 10), which con- authorities also promised to deal gener- stituted a more stringent form of lia ously with any unforeseen breaches of bility-side quantitative credit growth the agreements until fines were finally restriction. It also aimed at limiting lifted in October 1975. excessive credit growth, leverage and In 1955, the OeNB expanded the maturity mismatches. In view of the scope of the credit control agreements limited impact of minimum liquidity by including insurance companies and requirements, it limited banks’ credit instalment companies. The move was expansion to a specific percentage (e.g. motivated by the desire to avoid a shift- 70%) of the increase of their debt; this ing of credit supply to sectors outside was classified as a period-of-time ap- banking, i.e. shadow banking. Instal- proach (Zuwachsregelung) as the ratio ment companies had to reduce their out- referred to a change in debt over time.16 standing loans by 10% (OeNB, 1955, In case of a reduction of eligible debt, November 30). Insurance companies loans had to be reduced by the same were not allowed to grant loans anymore amount within an adequate time pe- and had to stop lending to credit instal- riod. Certain types of loans supportive ment institutions. From 1972 to 1981, of sustainable economic development they were again allowed to grant loans, were excluded from the credit ceiling, but only to a limited extent.19 which strengthened the qualitative As of 1957, capital-based limits on credit controls (e.g. export loans, agri- credit growth complemented the debt- cultural loans, investment loans and based limits in effect for the credit ceil- reconstruction loans). ing. In addition, the credit control The OeNB regularly monitored agreements with the different banking compliance with the quantitative provi- sectors were consolidated into one

16 As the bulk of bank debt consisted of bank deposits and savings accounts, the period-of-time approach was similar to a loan-to-deposit rule in flows rather than stocks. In contrast, the ratio under the point-in-time approach referred to the stock of credit and debt at a certain point in time. 17 In contrast, those that did not reach the credit ceiling were given additional room for providing credit supply (Krediterteilungsreserve). This stipulation was discontinued in 1955. 18 The fine for excessive credit supply beyond the limits of the credit ceiling corresponded to the bank rate times the amount of credit growth exceeding the limits. Besides the discount and lombard rate, the bank rate was another rate at which banks could refinance themselves with the OeNB. 19 Insurance companies were only allowed to increase loan supply by 6% of their credit stock as of November 30, 1972. The level, calculation base and reference period were changed several times.

172 OESTERREICHISCHE NATIONALBANK The changing role of macroprudential policy in Austria after World War II

agreement that aligned the minimum Six years later, in 1972, the OeNB liquidity requirements across banking pushed for a significant lowering of the sectors and effectively lowered the credit ceiling: it cut the loan-to-deposit credit ceiling by shifting from a period- ratio (point-in-time approach) by up to of-time to a point-in-time approach 12 percentage points to between 43% (Stichtagsregelung). This move limited and 68% and the ratio of loan growth the stock of outstanding loans to 45% to deposit growth (period-of-time ap- to 75% of the stock of debt plus 75% of proach) to between 35% and 37% as of the stock of bank capital.20 Again, cer- November 1972.21 The move was part tain types of loans (e.g. export loans, of a stability package which aimed at loans for agricultural purposes and in- slowing down the overheating econ- vestment loans, reconstruction loans) omy, excessive credit growth and a were exempted from these limits in substantial current account deficit support of a qualitative credit policy. (WIFO, 1972). The OeNB also re- In the following years, credit con- quired banks to curb consumer loan trol agreements were frequently tight- growth, to be more conservative in ex- ened (1962, 1965, 1972 and 1973) or tending real estate loans to residents loosened (1966, 1975). These changes and to refrain from advertising con- were components of more comprehen- sumer loans (Personalkredite) from sive stability packages that included fis- 1972 to 1975. In addition, banks agreed cal, income, monetary and capital ac- to refrain from selling domestic bonds count policy measures. While the and equity to nonresidents. However, credit ceiling in terms of debt and the the lowering of the credit ceiling re- corresponding period of time was mained ineffective and year-on-year changed frequently, the credit ceiling credit growth accelerated to about 20%. in terms of capital remained the same Therefore, another stability package in (75%) from 1957 until the termination 1973 lowered the credit ceiling further of the credit control agreements in by 10 percentage points and introduced 1981. For example, in 1966 the credit the Limes (see below). ceiling was raised by 2 percentage Contemporary assessments of the points in the context of somewhat various components of the credit con- muted credit expansion on the back of trol agreements differed. Bankers were poor investment and moderate eco- more in favor of qualitative credit con- nomic growth (WIFO, 1966a). In par- trols and minimum liquidity require- ticular, banks which easily complied ments, which were less intrusive. How- with the minimum liquidity require- ever, these measures were also less ments but had already reached the effective. In contrast, policymakers credit ceiling profited from this some- preferred the credit ceiling, which was what expansionary policy measure. both more restrictive and more effec- However, monetary policy was restrictive. This was the case because the tive at the same time, as e.g. the lom- agreements included specific limits for bard rate was raised to contain inflation each banking sector, so that differences pressures (WIFO, 1966b). in banks’ business structure could be

20 The latter share varied between banking sectors from 45% (head institution of savings banks) to 75% (smaller private banks, Volksbank credit cooperatives and head institution of Raiffeisen credit cooperatives). 21 The authorities also took further measures to contain regulatory arbitrage as banks substituted credit provision by granting excess credit lines. As of 1972, banks had to charge a fee of 0.5% on outstanding open credit lines.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 173 The changing role of macroprudential policy in Austria after World War II

taken into account. But they were also ticularly as deposit growth was strong more difficult to calibrate – particu- (Tichy, 1965). larly in times when the economic and Similarly, the stipulations of the financial environment was changing credit ceiling had only limited impact quickly (Tichy, 1965; Klauhs, 1963). on stemming credit growth. Until end- Also, the impact on the respective ob- 1972 banks did not exceed the credit jective could only be assessed with a lag ceiling and could have lent much more – which, however, is the case for any as the amount of granted loans was be- economic policy. Bankers feared that low the potential ceiling (with some the credit ceiling would distort compe- smaller exceptions within the savings tition, as banks with a history of loose bank sector (Neudörfer, 1968)). When credit policies were affected less (as credit growth became excessive in they started from higher initial values) 1972 despite the tightening of the than those with conservative policies. credit control agreements, the OeNB In addition, Austrian banks were at a convinced banks to accept the inclusion disadvantage compared with foreign in the credit control agreements of a banks that were not bound to the new type of credit policy, namely the credit control agreements (Klauhs, Limes (section 4.2.). 1964). Credit subsidies played an import- Altogether, credit control agreements ant role after 1945, but their interac- are found to have ambiguous impacts tion with the credit control agreements on credit growth in Austria between is difficult to assess.22 The OeNB found 1951 and 1981. It is not obvious whether that subsidized loans amounted to the respective objective of curbing or 38.5% of outstanding loans in 1977 strengthening credit growth was ob- (compared with 41.3% in 1965). About tained, particularly in light of the gen- 55% of the loan volume was granted to erous exemption clauses and wide- promote residential real estate develop- geared thresholds that were in place ment, 5% targeted agriculture and (Barfuss, 1975). This raises the ques- forestry and 36% focused on industrial tions of whether banks complied with and commercial investment or exports. the credit control agreements and Banks predominantly handled the lat- whether the specified thresholds were ter, while public sector entities focused tight enough. on the first two areas. This seems to be For most banks, the minimum in line with the objectives of qualitative liquidity requirements stipulated in the credit controls, which also prioritized credit control agreements were not industrial and commercial investment binding economically. Banks held large or exports. An assessment of the costs amounts of excess liquidity above and and benefits of these subsidies is beyond beyond the minimum requirements, and the scope of this paper. only very few banks ever had liquidity The era of credit control agree- shortfalls, so the requirements were ments ended in 1981, when the parties unlikely to have been very effective. involved could not agree on how to They were regarded too loose for the adapt the phased-out agreement of respective banks to effectively slow 1979. Credit policy fell out of favor down excessive credit expansion – par- with policymakers and central bankers 22 In 1965 and in 1977, the OeNB conducted two surveys on subsidized loans, which were defined as loans that can be granted below market rates because of interest rates subsidies, contributions to principal repayment or guarantees by public sector entities or institutions founded by the public sector for these purposes (OeNB, 1978).

174 OESTERREICHISCHE NATIONALBANK The changing role of macroprudential policy in Austria after World War II

alike (Greisinger, 1975; Elliot et al., volume. The Limes was modified sev- 2013). In December 1981, the OeNB’s eral times; it was loosened from 1% to General Council discussed the prolon- 1.3% per month in 1978, while it was gation and adaptation of the agreements tightened twice in 1980 to 1% and in line with the new legal foundations 0.5%, respectively, given strong credit that had been laid down in 1979 expansion. (OeNB, 1981). Ultimately, however, A special Limes was introduced in the OeNB and the Ministry of Finance the fall of 1977 to counteract strong decided to adopt a wait-and-see ap- growth in consumption. It reduced the proach. They would have agreed to a limit of consumer loan growth to prolongation, if – and only if – banks’ 0.55% per month (Chronik der Wäh unused additional lending capacity of rungspolitik). Consumption had increas- ATS 60 billion had been fully utilized. ingly shifted toward imported goods, Over the years, the liability-side credit which was a drag on the current ac- controls had not constituted economi- count. Economic and monetary policy- cally binding constraints for most makers saw a need to balance these banks. Given banks’ high excess lend- developments by supporting invest- ing capacity, the authorities did not ment, exports and tourism. Monetary expect a prolongation to be effective policy complemented the measure unless it entailed a substantial tighten- through preferential liquidity provision ing of the agreements. for these activities (WIFO, 1978; WIFO, 1981). 4.2 The Limes The Limes offered the advantage of The introduction of asset-side credit con- providing a very severe quantity re- trols in 1973,23 the so-called Limes, was striction with an immediate impact on a more intrusive measure of macropru- credit supply, but it had a number of dential policy. The Limes addressed ex- unintended side effects. First, the date cessive credit growth and leverage, and that served as a starting point for credit it was intended to surpass the low ef- growth in the following periods was fectiveness of the much less intrusive chosen arbitrarily. This rewarded banks liability-side measures. However, there that had expanded loans more aggres- was no legal framework for the Limes. sively before that date than more disci- The OeNB communicated a target plined banks, as the former were al- growth rate for the growth of credit to lowed to lend more after the Limes had nonbanks, which was 1% per month of entered into force. Second, measures the stock of loans extended to domestic that refer to a certain target date do not nonbanks as on November 30, 1972. take into account structural and sea- The OeNB incentivized compliance sonal developments in the market, with its target by excluding noncompli- which makes it harder for banks to deal ant banks from its refinancing opera- with seasonal variations in loan de- tions. In the following years, the refer- mand. Thus, especially bankers argued ence date was updated regularly, and it that such strict quantity measures referred to the end-of-year target stock should only be implemented for short of credit instead of the realized credit time periods (Klauhs, 1979).

23 The overarching objective of the Limes was to fight inflation by pursuing the intermediary objective of curtailing excessive credit growth. Other countries, e.g. the Netherlands, Sweden and Switzerland, had already introduced asset-side credit controls before 1973 (OeNB, 1969).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 175 The changing role of macroprudential policy in Austria after World War II

Chart 1 Credit-to-GDP gap in Austria from 1954 to 1984 % 25.0 1957: Credit control 1973: Limes 1977: Special 1980: 1981/1982: agreements consolidated implemented Limes implemented Limes Credit control 22.5 tightened agreements 20.0 and Limes end

17.5

15.0

12.5

10.0

7.5

5.0

2.5

0.0

–2.5 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 Credit-to-GDP gap Nominal GDP growth Nominal credit growth

Source: OeNB. Note: The credit-to-GDP gap is defined as the difference between the credit-to-GDP ratio and its long-term trend (calculated with a Hodrick-Prescott time series filter with a large smoothing parameter). The chart also shows the nominal credit and nominal GDP growth rates to better track whether the change of the credit-to-GDP gap was caused by a greater change in nominal credit or in nominal GDP growth rates. The ESRB recommends the credit-to-GDP gap as an early warning indicator for excessive unsustainable (real) credit growth and thus also for banking crises.

The Limes largely fulfilled the cial Limes was launched to limit con- OeNB’s high expectations. Asset-side sumer loan growth in 1977, nominal credit controls effectively limited credit credit growth dropped in Austria. The growth, even if banks had accumulated subsequent tightening of the Limes in excess liquidity and capital. To illus- 1980 was also successful in reducing trate the effectiveness of the Limes, we nominal credit growth, but the credit- look at nominal credit growth and the to-GDP gap widened until the abolish- credit-to-GDP gap in Austria between ment of the Limes in 1981/82 as nomi- 1954 and 1984 (chart 1). The latter is nal GDP growth decreased. an indicator of excessive credit growth When the Limes was implemented, recommended by the European Sys- the concept of a credit-to-GDP gap had temic Risk Board (ESRB, 2014). not been developed yet. However, there Chart 1 shows that the Limes was are some interesting insights we can effective in addressing its intermediary draw from its retrospective application. objective of mitigating excessive credit First, between 1954 and 1980, the growth in nominal terms. After its in- credit-to-GDP gap was below 2.5% – troduction in 1973, the credit-to-GDP the threshold defined by the Basel gap narrowed sharply due to a strong Committee of Banking Supervision decline in nominal credit growth, (BCBS, 2010) for excessive credit while nominal GDP growth remained growth (see section 5.3). But one has to relatively unchanged.24 When the spe- consider that the Limes and the credit

24 The slowdown of nominal GDP growth in 1974, after the first oil price shock, is likely to have contributed to the deceleration of nominal credit growth via a reduction of loan demand.

176 OESTERREICHISCHE NATIONALBANK The changing role of macroprudential policy in Austria after World War II

control agreements ultimately targeted (2007) argues that the gradual and inflation and therefore focused on nom- coordinated liberalization contributed inal excessive credit growth. Nominal to maintaining financial stability in credit growth rates of 22% at the time Austria, while abrupt liberalizations in the Limes was introduced justified pol- other countries contributed to twin icy action in light of the ultimate objec- crises (Kaminsky and Reinhart, 1999). tive of price stability (chart 1). The different target variables reflect the eman- 4.3 The Gentlemen’s Agreement cipation of macroprudential supervision As a complement to the credit control from monetary policy and the defini- agreements, the OeNB concluded a tion of different, though often interde- Gentlemen’s Agreement with the Aus- pendent objectives. Macroprudential trian banks in May 1971 to prevent supervision focuses on financial stabil- and/or sterilize speculative short-term ity and monetary policy on maintaining capital inflows.25 The Gentlemen’s price stability. Agreement was a voluntary agreement Finally, in 1981, the OeNB decided in which banks committed themselves to discontinue the general Limes for to depositing 40% of the value (in Aus- the sole purpose of removing the struc- trian schillings) of their capital inflows tural distortions that had resulted from in unremunerated accounts with the the application of the instrument over OeNB and to refrain from importing many years. The OeNB’s restrictive foreign capital to improve their liquid- monetary policy stance remained unaf- ity. The OeNB could have imposed fected. If the economic situation had higher minimum reserve requirements required it, the OeNB would immedi- unilaterally due to the amendment of ately have made full use of the instru- Article 43 Nationalbank Act in 1969. ment again (OeNB, 1980). However, The agreement aimed at mitigating the decision was made easier – like in risks arising from an excessive maturity the case of credit controls – by the ob- mismatch in the banking system and servation that the Limes had ceased to from foreign currency funding mis- be an economically binding constraint matches in the economy, as they pre- for banks. The OeNB discontinued the vented banks from providing longer- ban on the advertising of consumer term Austrian schilling-denominated loans and the special Limes in 1982. loans funded by short-term U.S. dollar- Demand for consumer loans was low denominated wholesale deposits. In and was expected to remain so in the August 1971, the OeNB briefly raised foreseeable future. In addition, credit that ratio from 40% to 100%. In policy fell out of favor due to a para- response to calls from exporters to digm change toward the greater liber- reduce the costs of hedging against ex- alization of financial markets, which change rate risk, the share was reduced had started in the late 1960s. The liber- again to 75% in September 1971 alization of the capital account and of (OeNB, 1971b). The agreements played the financial sector was expected to an important role in Austria’s exchange increase productivity and innovation in rate policy (Mooslechner et al., 2007; the real economy and the financial Handler, 2016; Schmitz, 2016). They sector (Braumann, 2002). Nowotny were prolonged frequently and comple-

25 Other countries such as Belgium, Denmark, Italy, the Netherlands and Switzerland introduced similar Gentlemen’s Agreements before 1971 (Schmitz, 1969).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 177 The changing role of macroprudential policy in Austria after World War II

mented the temporary reversal of the steps to again implement macropruden- liberalization of capital controls which tial policy measures to tackle specific took place between 1971 and 1976 financial stability issues. In Austria, (Mooslechner et al., 2007). The Gen- both measures to curb foreign currency tlemen’s Agreement was finally discon- lending and the 2012 sustainability tinued in 1980 (OeNB, 1980). package constitute predecessors to con- The Gentlemen’s Agreement (in temporary macroprudential measures. combination with capital controls) was The latter were a response to the ineffective in limiting foreign capital creasing connectivity and complexity inflows via banks, thereby mitigating of the financial system and a key lesson risks arising from maturity transforma- from the 2008 financial crisis. In additions in the financial system. The share tion, the single monetary policy fos- of new deposits that had to be held as tered the need for national macropru- unremunerated reserves at the OeNB dential policy as financial cycles were was very high, which drove a wedge be- not synchronized across the euro area tween the interest rates paid on depos- (Constâncio, 2015). In 2013, the ESRB its in Austria and those paid on deposits introduced five intermediary objectives by nonresidents in foreign currency of macroprudential policy as outlined (predominantly in U.S. dollars). The in section 1, which aim at supporting measure was very intrusive. Between the general objective of promoting 1971 (when the Gentlemen’s Agree- financial stability and reducing systemic ment was first introduced) and 1978, and procyclical risks to the financial short-term capital flows were con- system and the real economy (ESRB, tained quite well despite the strongly 2013). growing European dollar market and developments in the international ex- 5.1 Measures to curb foreign change rate system. We regard this as a currency lending contribution to avoiding twin crises in In 2003, the Austrian authorities turned Austria. Such concurrent banking and to macroprudential policy to curb for- currency crises were frequent in other eign currency lending. The related poli- European countries and especially cies aimed at mitigating or preventing Latin America. Twin crises often ac- excessive foreign currency credit growth companied a parallel liberalization of and excessive maturity mismatches and the capital account and a deregulation at limiting the concentration of direct of the banking sector. In many cases, foreign currency exposure with Austrian excessive maturity mismatches com- banks as well as the systemic impact of bined with foreign currency mis- misaligned incentives. matches due to volatile short-term cap- At the time, the portfolio of foreign ital flows were associated with twin currency loans held by Austrian banks crises.26 was large by international standards, which regulators identified as a poten- 5 Modern macroprudential tial source of systemic risk (Waschiczek, measures and their forerunners 2002; Auer et al., 2012). The FMA in Austria (which had replaced the Ministry of It took until 2003 for the Austrian Finance as banking supervisor in Austria authorities to take their first tentative in 2002) issued Minimum Standards,

26 See Kaminsky and Reinhart (1999) for data on twin crises.

178 OESTERREICHISCHE NATIONALBANK The changing role of macroprudential policy in Austria after World War II

which aimed at improving banks’ risk The measures to curb foreign cur- management practices concerning for- rency lending were not effective until eign currency lending and repayment 2008. The volume of foreign currency vehicle loans and at avoiding the exces- loans extended to Austrian households sive growth of foreign currency loans. increased constantly from EUR 18.3 In 2006, these measures were supple- billion in December 2003 to EUR 40 mented by information folders intended billion in October 2008.27 to improve the risk awareness of foreign Only the more intrusive supervi- currency borrowers. However, these sory efforts in 2008, 2010 and 2013 legally nonbinding soft-law measures proved to be effective, as the share of ex- failed to achieve the intended goal, as change rate-adjusted foreign currency neither banks nor borrowers fully loans extended by Austrian banks to grasped the risks related to foreign cur- customers in Austria and Central, East- rency lending (refinancing risk, concen- ern and Southeastern Europe (CESEE) tration risk and political risk for banks; decreased significantly thereafter. Still, exchange rate risks, etc. for borrowers). also the financial crisis of 2008 (includ- As the global financial crisis led to ing the appreciation of the Swiss franc heightened risk awareness, the FMA against the euro and the losses associ- strongly recommended in 2008 that the ated with repayment vehicles) is likely banking sector should no longer extend to have had an impact on the demand foreign currency loans to households. for foreign currency loans, which The latest measures were the 2013 makes it difficult to assess the effective- FMA Minimum Standards (FMA, ness of the supervisory measures. In 2013a, b) which integrated the ESRB absolute terms, foreign currency loans recommendations concerning the grant- to Austrian households declined from ing of foreign currency loans (ESRB, EUR 40 billion in October 2008 to less 2011). The new recommendation effec- than EUR 24 billion at the end of 2015. tively limited the issuance of euro-de- In exchange rate-adjusted terms, foreign nominated consumer and mortgage currency loans and particularly Swiss loans to those households and SMEs that franc-denominated loans extended to were effectively hedged (e.g. via eu- Austrian households decreased by over ro-denominated income or revenue). In 50% over the same period (OeNB, addition, Austrian banks were discour- 2015b). Since mid-2014, new foreign aged from granting new repayment vehi- currency loans supplied by CESEE sub- cle loans denominated in foreign cur- sidiaries have been almost entirely de- rency. While the Minimum Standards of nominated in euro in compliance with 2008 and 2013 are legally nonbinding, macroprudential policy provisions. they have increased banks’ legal risks, and subsequently costs, associated with 5.2 The sustainability package these products. The FMA expects credit In 2012, the FMA and the OeNB intro- institutions to comply with the restric- duced a sustainability package (FMA, tive interpretation of the Minimum 2012) to address potential contagion Standards with regard to banks’ legal effects emanating from Austrian banks’ due diligence obligations (Article 39 CESEE subsidiaries and to improve the Austrian Banking Act). resilience of Austrian banks’ business

27 The same trend was observed for loans to nonbank financial intermediaries (+EUR 1.6 billion) but not for loans to nonfinancial corporations (–EUR 10 billion).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 179 The changing role of macroprudential policy in Austria after World War II

models. From today’s perspective, its The sustainability package has been intermediate objectives were mitigat- effective in improving the local refi- ing or preventing excessive maturity nancing structure of Austrian banks’ mismatches. subsidiaries in CESEE. Their loan-to- The package covered Austria’s three deposit ratio decreased from 117% in largest banks, Erste Group, Raiffeisen 2008 to 90% in the third quarter of Zentralbank Österreich AG and Uni- 2015 due to the strong growth (by over Credit Bank Austria, which committed 30%) of savings deposits of local non- in a voluntary agreement to frontloading banks in CESEE. Loan growth is now the Basel III common equity tier 1 considered to be more sustainable as it (CET1) requirement of holding 4.5% of is mainly financed on a local basis. risk-weighted assets as minimum capital However, the contribution of the sus- requirements at the consolidated level tainability package to improving Aus- by January 2013. In addition, the three trian banks’ CESEE subsidiaries’ busi- banks also agreed to frontload the capi- ness models is hard to assess, as the tal conservation buffer of 2.5% CET1 of financial crisis has led to a substantial risk-weighted assets so that their target decrease of current account imbal- CET1 ratio amounted to 7%. The agree- ances, and thus of capital imports, in ment also included a monitoring tool CESEE. Loan demand has also declined focused on the local funding capacity of as a result of the financial crisis. Austrian banks’ CESEE subsidiaries. The OeNB closely monitors this loan-to- 5.3 Modern macroprudential local stable funding ratio (LLSFR) – the measures ratio of local loans in CESEE to local In contrast to their predecessors, mod- deposits and other sources of stable ern macroprudential measures are em- funding – to reduce the excessive bedded in the comprehensive legal and buildup of foreign currency mismatches institutional framework of macropru- in CESEE and to avoid excessive loan dential supervision that was introduced growth at these three banks. in 2014 in Austria and the EU. In According to the Austrian super Austria, all relevant national financial visors’ ex ante analysis, Austrian banks’ stability stakeholders are represented subsidiaries which before the recent on the Financial Market Stability financial crisis had recorded high loan Board28 (FMSB) (for details, see Eiden- growth that was not backed by strong berger et al., 2014). The FMSB may local stable funding (which translates issue recommendations to the FMA, into stock and flow LLSFR exceeding which is the designated national macro- 110%) were more likely to exhibit prudential authority. The OeNB plays a higher loan loss provisioning rates (due pivotal role within the Austrian frame- to their higher vulnerability to credit work for macroprudential supervision: risks and write-offs) during the ensuing it is responsible for identifying potential crisis. Thus, to improve the refinancing systemic risks and for providing the an- structure of selected subsidiaries, the alytical underpinning of macropruden- Austrian supervisors agreed on using tial measures (including impact assess- the LLSFR as a monitoring tool and an ments of policy measures). In addition, early warning indicator for nonsustain- it provides the secretariat to the FMSB. able, excessive credit growth.

28 The activities and deliberations of the FMSB are documented on its website (www.fmsg.at).

180 OESTERREICHISCHE NATIONALBANK The changing role of macroprudential policy in Austria after World War II

In 2015, the FMSB issued a recom- Also in 2016, the countercyclical cap- mendation to the FMA as Austria’s ital buffer (CCyB) framework entered macroprudential authority to activate into force. It aims at mitigating or pre- the systemic risk buffer (SRB) to protect venting excessive credit growth and the Austrian financial system and the leverage. Until its most current meet- Austrian economy against long-term ing on June 1, 2016, the FMSB recom- noncyclical systemic risks.29 The FMSB mended setting the CCyB rate at 0% of advocated the SRB as the Austrian risk-weighted assets. The CCyB aims at banking sector was exposed to a spe- smoothing the pronounced cyclicality cific combination of systemic risks aris- in the financial system by implementing from the above-average size of its ing a buffer in a phase of excessive banking sector in a European compari- credit growth which is then released in son, a high exposure to emerging econ- a downturn to absorb loan losses with- omies in Europe, insufficient prepara- out impeding banks’ access to funding tion for the reduction (or removal) of markets. As such, it aims at reducing the implicit state guarantee, relatively the procyclical effects of asset write- low capital ratios and specific owner- downs, loan losses and market expecta- ship structures (OeNB, 2015a; FMSB, tions regarding bank capital ratios 2015c). (FMSB, 2015b; FMSB, 2016). In June 2016, the rules on the buffer In 2016, the FMSB discussed the for other systemically significant institu- need for additional macroprudential tions (O-SII buffer) entered into force.30 instruments in line with the ESRB Of the SRB and the O-SII buffer, only recommendation (ESRB, 2013). The the higher one is binding. All banks that OeNB initiated the discussion and has are subject to the O-SII buffer are also advocated the extension of the macro- subject to a higher SRB, so that the for- prudential toolkit by borrower-based mer does not increase their CET1 ra- (asset-side) instruments, such as limits tios. The O-SII buffer aims at reducing on loan-to-value-ratios (LTVs), debt-to- the systemic risk stemming from the income ratios (DTIs), and debt service- failure of a systemically significant in- to-income ratios (DSTIs). These mac- stitution. roprudential real estate instruments The capital conservation buffer (CCB) are expected to contribute to dampen- aims at shielding banks’ minimum cap- ing credit growth during the upswing ital requirements from bank losses. In of the credit cycle by moderating the good times banks accumulate a buffer credit cycle and thereby strengthening which they can draw on under stress to banks’ and borrowers’ resilience. Con- absorb losses without endangering their sequently, the FMSB suggested in Feb- viability. The respective requirement ruary 2016 to establish the legal basis entered into force in 2016 and is phased for these essential macroprudential in- in until 2019 when it will reach 2.5% of struments (FMSB, 2016). CET1 capital relative to risk-weighted There are not yet any ex post stud- assets. ies on the effectiveness of the SRB as it

29 The SRB entered into force on January 1, 2016, with a phase-in period until 2019. It amounts to up to 2% of risk-weighted assets held in CET1, in addition to the minimum capital requirement of 4.5% of risk-weighted assets in CET1 under Pillar 1 plus the applicable Supervisory Review and Evaluation Process (SREP) ratio (FMSB, 2015a; FMSB, 2015b). 30 A similar buffer applies to globally systemically important institutions. However, no such institutions are domiciled in Austria.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 181 The changing role of macroprudential policy in Austria after World War II

was introduced only on January 1, ror learning process. Policymakers’ ul- 2016, for selected Austrian banks; timate objective was to maintain price however, the OeNB conducted a com- stability by addressing the intermediate prehensive ex ante impact assessment, objectives of mitigating or preventing which estimates the SRB’s net impact excessive credit growth, leverage and on the economy in Austria by compar- maturity mismatches. There was no ing its costs and benefits and finds that sound legal basis for macroprudential the activation of the SRB would have a policy in place in that period; Austrian positive net effect on the economy. The authorities had to rely on voluntary SRB increases banks’ weighted average agreements with the banking industry. cost of capital as it induces banks to The Kreditwesengesetz (KWG) of substitute debt capital by more expen- 1979 eventually introduced a legal sive equity capital. The analysis assumes basis, but macroprudential policy mea- that banks would pass on these addi- sures (e.g. credit control agreements) tional opportunity costs to borrowers fell out of favor with policymakers and by increasing the interest rate on new central bankers at that time as capital loans to nonbanks (Kopp et al., 2010). account liberalization and the deregula- This would lead to an average widening tion and internationalization of finance of credit spreads by a mere 0.08 basis gained momentum – a period we refer points in the medium term (between to as the second phase (from 1982 to 0.05 basis points and 0.13 basis points 2014). Austrian authorities never made depending on the target return on use of their new legal macroprudential equity (ROE) of 10% to 15%). Conse- powers, which were abolished in 1986, quently, the additional SRB-related only seven years after their introduc- capital requirement for 2017 would re- tion. It took another 28 years until sult in a minor increase of credit costs macroprudential supervision was for- for the real economy which would, in mally introduced in Austria in 2014 turn, result in rather insignificant nega- (third phase from 2014 to present). tive short- and medium-term gross This time, financial stability as distinct effects on GDP growth (below 0.01% from price stability has been the ulti- of GDP accumulated from 2016 to mate objective of macroprudential su- 2018). In terms of long-term benefits, pervision. the OeNB analysis finds that these out- Looking at the employed tools in weigh the costs of the SRB activation, greater detail, the first credit control as it would reduce the probability and agreements in 1951 included both quali- the social costs of banking crises in tative and quantitative credit controls. Austria. The overall net impact on the Initially, the latter comprised minimum economy in Austria is therefore found liquidity requirements only; over time, to be positive. these were complemented by capital- based limits on credit growth, credit 6 Conclusions ceilings (liability-side credit controls) We find that macroprudential policy in and the Limes (asset-side credit con- Austria has a long and changeful history trols). The measures were implemented that started as early as in 1951. It in the form of voluntary agreements be- evolved in three phases. The first phase, tween banks and policymakers as there from 1951 to 1982, was dominated by was no legal basis for macroprudential an expansion of the macroprudential policy in place in Austria at the time. toolset based on an active trial-and-er- The liability-side credit controls were

182 OESTERREICHISCHE NATIONALBANK The changing role of macroprudential policy in Austria after World War II

nonintrusive measures; they hardly ever systemic risk buffer was designed to ef- constituted economically binding con- fectively reduce the systemic risk po- straints, and their impact on credit tentially associated with the operations growth was low. However, asset-side of Austrian banks eligible for the sys- credit controls (the Limes agreements) temic risk buffer by improving their were quite intrusive and effective in capitalization. curtailing excessive loan growth. This analysis identifies key criteria In 1971, the Gentlemen’s Agree- for the effectiveness of macroprudential ment was introduced with the aim of policy: first, macroprudential policy re- curtailing short-term capital inflows. It quires a sound legal basis, and second, was very intrusive and effective at the macroprudential policy measures have same time. to be quite intrusive to effectively cur- In 2014, a sound legal basis for mac- tail the buildup of systemic risk. Less roprudential supervision was intro- intrusive measures become effective duced – a key lesson from the 2008 above all by increasing the shock-ab- financial crisis. In 2016, the first legally sorbing capacity of the financial system binding macroprudential measures since once risks materialize. 1945 entered into force in Austria. The

References Auer, R. A., S. Kraenzlin and D. Liebeg. 2012. How Do Austrian Banks Fund Their Swiss Franc Exposure? In: Financial Stability Report 24. 54–61. Austrian Banking Act – Bundesgesetz über das Bankwesen (BWG). 1993. Federal Law Gazette No. 532/1993. Austrian Institute of Economic Research (WIFO). 1966a. Monatsberichte des Öster reichischen Instituts für Wirtschaftsforschung 39 (5). Austrian Institute of Economic Research (WIFO). 1966b. Monatsberichte des Öster reichischen Instituts für Wirtschaftsforschung 39 (7). Austrian Institute of Economic Research (WIFO). 1972. Monatsberichte des Öster reichischen Instituts für Wirtschaftsforschung 45 (12). Austrian Institute of Economic Research (WIFO). 1978. Monatsberichte des Öster reichischen Instituts für Wirtschaftsforschung 51 (3). Austrian Institute of Economic Research (WIFO). 1981. Monatsberichte des Öster reichischen Instituts für Wirtschaftsforschung 54 (3). Barfuss, H. 1975. Die Entwicklung der Kreditkontrolle in Österreich. In: Theorie und Praxis der Kreditkontrolle. Schriftenreihe des Österreichischen Forschungsinstituts für Sparkassenwesen 15(1). 55–73. Basel Committee on Banking Supervision (BCBS). 2010. Countercyclical capital buffer proposal. www.bis.org/publ/bcbs172.pdf (retrieved on July 12, 2016). Braumann, B. 2002. Financial Liberalization in Austria: Why so Smooth? In: Financial Stability Report 4. 100–114. Chronik der Währungspolitik. Various years. (Chronology of monetary policy). OeNB Bank History Archives . Clement, P. 2010. The term “macroprudential”: origins and evolution. In: BIS Quarterly Review. March. 59–67.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 183 The changing role of macroprudential policy in Austria after World War II

Constâncio, V. 2015. Financial stability risks, monetary policy and the need for macro-prudential policy. Speech by Vítor Constâncio. Vice-President of the ECB. Warwick Economics Summit. February 13. Dudley, W. C. 2015. Is the Active Use of Macroprudential Tools Institutionally Realistic? Panel remarks at the Macroprudential Monetary Policy Conference. Federal Reserve Bank of Boston. Boston, MA. ECB. 2011. The implementation of monetary policy in the Euro. General documentation on Euro monetary policy instruments and procedures. www.ecb.europa.eu/pub/pdf/other/gendoc 201109en.pdf?03fdeef26c45121ee9e0212a39517f08 (retrieved on July 13, 2016). Eidenberger, J., D. Liebeg, S. W. Schmitz, R. Seliger, M. Sigmund, K. Steiner, P. Strobl and E. Ubl. 2014. Macroprudential Supervision: A Key Lesson from the Financial Crisis. In: Financial Stability Report 27. 83–94. Elliott, D., Feldberg, G. and A. Lehnert. 2013. The History of Cyclical Macroprudential Policy in the United States. Staff working papers in Finance and economics Discussion Series (Federal Reserve Board, Washington, D.C.). www.federalreserve.gov/pubs/feds/2013/201329/ 201329pap.pdf (retrieved on August 8, 2016). ESRB. 2011. Recommendation of the European Systemic Risk Board of 21 September 2011 on lending in foreign currencies (ESRB/2011/1). Frankfurt. www.esrb.europa.eu/pub/pdf/recommendations/2011/ESRB_2011_1.en.pdf (retrieved on July 28, 2016). ESRB. 2013. Recommendation of the European Systemic Risk Board on intermediate objectives and instruments of macro-prudential policy (ESRB/2013/1). Frankfurt. www.esrb.europa.eu/ pub/pdf/recommendations/2013/ESRB_2013_1.en.pdf (retrieved on July 13, 2016). ESRB. 2014. Recommendation of the European Systemic Risk Board on guidance for setting countercyclical buffer rates (ESRB/2014/1). Frankfurt. www.esrb.europa.eu/pub/pdf/recommendations/2014/140630_ESRB_Recommendation.en. pdf?83075f19bd8f21d8a3b8e6afe7bea49b (retrieved on July 13, 2016). Federal Act on the Institution and Organisation of the Financial Market Authority. Federal Law Gazette I No. 97/2001 as amended in Federal Law Gazette I No. 77/2011. Financial Market Authority (FMA). 2012. Supervisory guidance on the strengthening of the sustainability of the business models of large internationally active Austrian banks. March 14. Financial Market Authority (FMA). 2013a. Foreign Currency Loans and Loans with Repayment Vehicles. Vienna. Financial Market Authority (FMA). 2013b. FMA-Minimum Standards for the Risk Manage- ment and Granting of Foreign Currency Loans and Loans with Repayment Vehicles (FMA- FXTT-MS). Vienna. Financial Market Stability Board (FMSB). 2015a. Fourth meeting of the Financial Market Stability Board. Press releases. fmsg.at/en/publications/press-releases/fourth-meeting.html (retrieved on July 13, 2016). Financial Market Stability Board (FMSB). 2015b. Fifth meeting of the Financial Market Stability Board. Press releases. fmsg.at/en/publications/press-releases/fifth-meeting.html (retrieved on July 13, 2016). Financial Market Stability Board (FMSB). 2015c. Adjustment of the recommendation issued after the fourth meeting of the Financial Stability Market Board regarding the activation of the systemic risk buffer. fmsg.at/en/publications/warnings-and-recommendations/recommendation-4.html (retrieved on July 13, 2016). Financial Market Stability Board (FMSB). 2015d. Sixth meeting of the Financial Market Stability Board. Press releases. fmsg.at/en/publications/press-releases/sixth-meeting.html (retrieved on July 13, 2016).

184 OESTERREICHISCHE NATIONALBANK The changing role of macroprudential policy in Austria after World War II

Financial Market Stability Board (FMSB). 2016. Seventh meeting of the Financial Market Stability Board. Press releases. fmsg.at/en/publications/press-releases/seventh-meeting.html (retrieved on July 13, 2016). Financial Stability Board, BIS and IMF. 2011. Macroprudential policy tools and frameworks: Progress report to G20. October. www.fsb.org/wp-content/uploads/r_111027b.pdf?page_ moved=1 (retrieved on July 13, 2016). Foreign Exchange Act (Devisengesetz). 1946. Federal Law Gazette No. 162/1946. Galati, G. and Moessner, R. 2011. Macroprudential policy – a literature review. BIS Working Papers No. 337. Gesetz zur Neuordnung der wirtschaftlichen und finanziellen Lage der Kreditun- ternehmen (Reconstruction Act). 1955. Federal Law Gazette No. 183/1955. Goodhart, C. 1989. Money, Information and Uncertainty. MacMillan. London. Greisinger, H. 1975. Zusammenfassung. In: Theorie und Praxis der Kreditkontrolle. Schriften- reihe des Österreichischen Forschungsinstituts für Sparkassenwesen. 15 (1). 103–106. Habermeier, K., A. Kokenyne and C. Baba. 2011. The Effectiveness of Capital Controls and Prudential Policies in Managing Large Inflows. IMF Staff Discussion Note SDN/11/14. Heller, H. 1980. Das Zusammenwirken von Notenbank, Finanzministerium und Kreditapparat nach dem KWG 1979. In: Notenbank und Kreditapparat. Schriftenreihe der Österreichischen Bankwissenschaftlichen Gesellschaft LVIII/1980. 32–44. Kaminsky, G. and C. Reinhart. 1999. The Twin Crises: The Causes of Banking and Balance- of-Payments Problems. The American Economic Review 89 (3). 473–500. Kelber, A., and E. Monnet. 2014. Macroprudential policy and quantitative instruments: a Euro- pean historical perspective. In: Financial Stability Review 18. Banque de France. 151–160. Klauhs, H. 1963. Möglichkeiten und Grenzen einer differenzierten Kreditpolitik. Österreichisches Bankarchiv. 206–214. Klauhs, H. 1964. Die Kreditkontrollabkommen im Lichte der Bankpolitik. Österreichisches Bankarchiv. 34–45. Klauhs, H. 1971. Mindestreserve und Kreditkontrollabkommen als Instrumente der öster reichischen Konjunkturpolitik. In: Schmitz, W. (ed.). Geldwertstabilität und Wirtschaftswachs- tum. Springer. Vienna. New York. 125–149. Klauhs, H. 1979. Die „indirekte“ Kreditrestriktion – Weg oder Irrweg? Österreichisches Bank archiv I. 19–33. Kopp, E., C. Ragacs and S. W. Schmitz. 2010. The Economic Impact of Measures Aimed at Strengthening Bank Resilience – Estimates for Austria. In: Financial Stability Report 20. 86–114. Kreditlenkungsgesetz – Credit steering act. 1945. State Law Gazette No. 43/1945. July 4. Kreditwesengesetz (KWG) – Austrian credit services act. 1979. Federal Law Gazette No. 63/1979. February 20. Linhardt, H. 1963. Der unlösliche Widerspruch zwischen der qualitativen Kreditkontrolle der Geschäftsbanken und den quantitativen Kreditnormen der staatlichen Bankenaufsicht. Öster reichisches Bankarchiv. 2–18. Ministry of Finance. 1955. Entwurf eines Bundesgesetzes über das Kreditwesen. Austrian State Archives. Reference no. 85.154 15-A/55. Ministry of Finance. 1969. Ministeratsvortrag betreffend den Entwurf des Kreditwesengesetzes 1969. Austrian State Archives. Reference no. 311.396 15 b/69. Monnet, E. 2015. Monetary Policy without Interest Rates: Evidence from France's Golden Age (1948 to 1973) Using a Narrative Approach. In: American Economic Journal: Macroeconomics 5(4). 137–169.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 185 The changing role of macroprudential policy in Austria after World War II

Mooslechner, P., S. W. Schmitz and H. Schubert. 2007. From Bretton Woods to the Euro: the Evolution of Austrian Monetary Policy from 1969 to 1999. In: From Bretton Woods to the Euro – Austria on the Road to European Integration. Workshops – Proceedings of OeNB Workshops 11. 21–44. Nationalbank Act (NBG). 1955. Federal Law Gazette No. 184/1955. Nationalrat. 1955. 611 der Beilagen zu den stenographischen Protokollen des Nationalrates. VII. Gesetzgebungsperiode. Nationalrat. 1979. 844 der Beilagen zu den stenographischen Protokollen des Nationalrates. XIV. Gesetzgebungsperiode. Nationalrat. 1986. 934 der Beilagen zu den stenographischen Protokollen des Nationalrates. XVI. Gesetzgebungsperiode. Nowotny, E. 2007. Die Hartwährungspolitik und die Liberalisierung des Kapitalverkehrs sowie des Finanzsektors. In: From Bretton Woods to the Euro – Austria on the Road to European Integration. Workshops – Proceedings of OeNB Workshops 11. 45–68. OeNB. 1951–1983. Mitteilungen der Oesterreichischen Nationalbank (Communications of the Oesterreichische Nationalbank). Monthly issues. Vienna. OeNB. 1968. Protokoll des Unterausschusses für Offenmarktoperationen und Mindestreserve- politik (Minutes of the subcommittee on open market operations and minimum reserve policy). OeNB Bank History Archives 25B. 1968. 612 PAD/DirAkt/AktNr. 1–46. OeNB. 1969. Protokolle der Sitzungen des Generalrats (Minutes of the meetings of the General Council). No. 257. OeNB Bank History Archives. OeNB. 1971a. Aktennotiz über die Sitzung des Unterausschusses des Generalrates für Währung- spolitik am 8. September 1971 (Memo of the meeting of the Subcommittee for Monetary Policy of the General Council on September 8). OeNB Bank History Archives. OeNB. 1971b. 276. Generalratssitzung am 15. September 1971. Anlage 4. Währungspolitische Maßnahmen (276th meeting of the General Council on September 15. Annex 4. Monetary policy measures). OeNB Bank History Archives. OeNB. 1978. Sondererhebung über subventionierte Kredite. Berichte und Studien. 427–438. OeNB. 1980. Protokolle der Direktoriumssitzungen (Minutes of the meetings of the Governing Board). No. 1213. OeNB Bank History Archives. OeNB. 1981. Protokolle der Direktoriumssitzungen (Minutes of the meetings of the Governing Board). No. 1223. OeNB Bank History Archives. OeNB. 2015a. Financial Stability Report 29. Vienna. OeNB. 2015b. Financial Stability Report 30. Vienna. Randa, G. 1975. Gedanken zur Problematik der Kreditkontrolle. In: Theorie und Praxis der Kreditkontrolle. Schriftenreihe des Österreichischen Forschungsinstituts für Sparkassenwesen 15 (1) 74–83. Reichskreditwesengesetz (RKWG) – Reich Banking Act. 1938. Kundmachung des Reichsstatthalters in Österreich, wodurch die Verordnung zur Einführung von Gesetzen über das Kredit- und Zahlungswesen im Lande Österreich vom 1. Oktober 1938 bekannt gemacht wird. German Reich Law Gazette I S 1329/1938. Reichskreditwesengesetz (RKWG) – Reich Banking Act. 1939. Reichsgesetz über das Kreditwesen. September 25. German Reich Law Gazette I S 1955/1939. Scheithauer, M. 1975. Währungspolitik und Kreditrestriktionen: In: Theorie und Praxis der Kreditkontrolle. Schriftenreihe des Österreichischen Forschungsinstituts für Sparkassenwesen 15 (1). 49–55.

186 OESTERREICHISCHE NATIONALBANK The changing role of macroprudential policy in Austria after World War II

Schmitz, W. 1969. Balance of payments and the creation of money – Influence of the net balance of payments on changes in money supply. Lecture given at Harvard University, Cambridge, MA. October 6. OeNB Bank History Archives I/06.Sc. Seidel, H. 2005. Österreichs Wirtschaft und Wirtschaftspolitik nach dem Zweiten Weltkrieg. Manz. Vienna. Solow, R. M. 1956. A contribution to the theory of economic growth. In: Quarterly Journal of Economics 70 (1). 65–94. Swan, T. W. 1956. Economic growth and capital accumulation. In: Economic Record 32 (2). 334–361. Tichy, G. 1965. Die Kreditkontrollabkommen. In: Wirtschaftspolitische Blätter 12 (3/4).153–157. Waschiczek, W. 2002. Foreign Currency Loans in Austria – Efficiency and Risk Considerations. In: Financial Stability Report 4. 83–99.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 187 The changing role of macroprudential policy in Austria after World War II Table 1 Table Targeted banking sector Targeted Step-by-step inclusion of other banking sectors until 1960s except for for inclusion of other banking sectorsStep-by-step until 1960s except (OeKB) Kontrollbank banks, e.g. Oesterreichische specialized All banking sectors except specialized banks, e.g. OeKB specialized All banking sectors except All banking sectors except specialized banks, e.g. OeKB specialized All banking sectors except Targeted banking sector Targeted Step-by-step inclusion of other banking sectors until 1960s except for for inclusion of other banking sectorsStep-by-step until 1960s except banks, e.g. OeKB specialized All banking sectors except specialized banks, e.g. OeKB specialized All banking sectors except Targeted banking sector Targeted Joint stock inclusion of banks → step-by-step banks and smaller private Joint banks, e.g. OeKB specialized other banking sectors except Insurance companies Instalment companies All banking sectors except specialized banks, e.g. OeKB) specialized All banking sectors except All banking sectors except specialized banks, e.g. OeKB specialized All banking sectors except All banking sectors except specialized banks, e.g. OeKB specialized All banking sectors except Insurance companies Insurance All banking sectors except specialized banks, e.g. OeKB banks, e.g. specialized except sectors banking All All banking sectors except specialized banks, e.g. OeKB banks, e.g. specialized except sectors banking All All banking sectors except specialized banks, e.g. OeKB banks, e.g. specialized except sectors banking All % of liabilities % to 75 % of equity and 45 % reduction of the credit stock and strict quality checks for new loans – Credit supply only for sustainable purposes for only supply – Credit Contain inflation support sustainable domestic export in the economy, investment, the misallocation of credit and tourism Avoid the Ministry with the OeNB) and banking sectors between agreements of Finance (in accordance control joint credit Voluntary Provision – Provisions remained in place during the consolidation of credit control agreements control during the consolidation of credit in place remained – Provisions – Ban on advertising of consumer loans until suspension in 1982 – Suspension Contain inflation and avoid tensions in the foreign exchange market exchange in the foreign tensions Contain inflation and avoid Avoid excessive credit growth and excessive maturity transformation in the banking systemVoluntary joint credit control agreements between the Ministry of Finance in accordanceProvision with the OeNB and the banking sectors – 10% of liabilities to be held as first-grade and 20% as second-grade liquidity – Specific provisions (e.g. the percentage share) varied across banking sectors – Consolidation of credit control agreements: of liabilities 15% to be held as first-grade and 25% as second-grade liquidity Contain inflation, avoid tensions in the foreign exchange market and limit regulatory arbitrage by insurance and instalment and limit regulatory arbitrage by companies market exchange in the foreign tensions Contain inflation, avoid in the banking system maturity transformation and excessive growth credit excessive Avoid the Ministry with the OeNB) and banking sectors between agreements of Finance (in accordance control joint credit Voluntary Provision – Credit growth restricted to 70% of banks’ liability increase (excluding equity) – Bank-specific ratio tightened to 50% of banks’ liability increase (excluding equity) companies insurance by supply credit of – Prohibition – 10 – Consolidation of credit control agreements: credit stock restricted to 75 – Same as in 1951 – Lowering of the credit ceiling by 3 percentage points – Monthly credit growth restricted to to 72% 43% of liability increase – Credit ceiling raised by 2 percentage points – Credit ceiling lowered by 5–7 percentage points – Monthly credit growth restricted to between and 35% 37% of liability increase – Credit stock may be expanded by only 6% – Credit ceiling lowered percentage by 10 points – Credit ceiling raised percentage by 10 points – Suspension Credit ceiling (Kreditplafond) controls) (liability-side quantitative credit 1951/04 to 1981/06 1951/04 to Qualitative credit controls Qualitative credit Ultimate objective Intermediate objectiveIntermediate Legal basis Duration 1957/04 to suspension 1957/04 to 1982/09 1972/09 to 1981/06 Minimum liquidity requirements (liability-side quantitative credit controls) (liability-side quantitative credit liquidity requirements Minimum objective Ultimate objective Intermediate basis Legal Duration 1952/01 1957/04 to suspension1957/04 1981/06 Ultimate objective Ultimate objective Intermediate basis Legal Duration 1951/10 to major 1951/10 amendment 1952/04 to major amendment to major1955/11 amendment 1955/12 to major1955/12 amendment to major1957/04 amendment 1962/08 to major amendment 1966/07 to major amendment 1972/10-12 to1972/10-12 major amendment 1972/12 to major1972/12 amendment 1973 to major1973 amendment 1975 to suspension1975 1981/06 Annex: Milestones of macroprudential policy in Austria since the 1950s

188 OESTERREICHISCHE NATIONALBANK The changing role of macroprudential policy in Austria after World War II Table 1 continued 1 Table All banking sectors except specialized banks, e.g. OeKB specialized All banking sectors except OeKB banks, e.g. specialized except sectors banking All banks, e.g. OeKB specialized All banking sectors except Targeted banking sector Targeted All banking sectors except specialized banks, e.g. OeKB specialized All banking sectors except All banking sectors except specialized banks, e.g. OeKB banks, e.g. specialized except sectors banking All Targeted banking sector Targeted All Austrian banks Targeted banking sector Targeted All Austrian banks All Austrian banks and in particular those operating in CESEE All Austrian banks All Austrian banks Targeted banking sector Targeted AG, Zentralbank Österreich Raiffeisen 3 Austrian banks: Erste Group, Bank Austria UniCredit Targeted banking sector Targeted Any banks’ loan exposure located in Austria located banks’ loan exposure Any 12 Austrian12 banks 7 Austrian banks %; in September 1971: in September 1971: %; % % phasing-in 2016–19 % for subsidiaries% for % per month, phasing-in until 1978 % per month % %, phasing-in 2016–19 % of foreign capital inflows on unremunerated accounts at the OeNB (August 1971: rate was raised to 100 on unremunerated accounts at the OeNB (August 1971: capital inflows % of foreign % and tightening of the special Limes to 0.35 of the special Limes to % and tightening %) – Special Limes: consumer loan growth limited to 0.55 to limited – Special Limes: consumer loan growth 0.5 of the Limes to – Tightening – Suspension of the Limes – Growth of credit to nonbanks limited to 1 to nonbanks limited to of credit – Growth Contain inflation consumption) (in particular in the economy credit-financed much and misallocation too of credit growth credit Contain excessive the OeNB guidance published by Policy Provision Suspension of the special Limes special the of – Suspension rate was lowered to 75 rate was lowered Avoid excessive build-up of foreign capital inflows and excessive maturity transformation capital inflows and excessive build-up of foreign excessive Avoid support exports and capital inflows short-term speculative sterilize and/or Prevent Voluntary agreements between the OeNB and Austrian banks Provision Banks committed to depositing 40 growth credit currency Contain foreign loans of Austrian vehicle banks and repayment currency risk management practices of foreign Improve Guiding principles of the FMA and OeNB Provision FMA Minimum Standards (improvement of risk management practices) Standards (improvement FMA Minimum Extension of FMA Minimum Standards Extension of FMA Minimum New FMA Minimum Standards New FMA Minimum Publication of information foldersPublication currency on foreign risks Strengthen the shock-absorbing capacity of Austrian banks’ subsidiaries operating in CESEE the shock-absorbing capacity Strengthen Avoid a dry-up of liquidity and improve refinancing options at the subsidiary level as wellGuiding as strengthen principles of banking the FMA and the groups’ OeNB capital adequacy Provision Full implementation of the Basel III rules in respect of CET1 capital at the consolidated level Loan-to-local stable funding ratio 100 Loan-to-local stable (LLSFR) of below Strengthen the risk-bearing capacity of banks and financial stability of banks and financial the risk-bearing capacity Strengthen Contain excessive credit growth address (CCB), systemic vulnerability and systemicEU Capital cluster risk Requirements (SRB) and idiosyncratic Regulation banking (CRR) and Austrian risks Banking for the system Act (O-SIIB) (BankwesengesetzProvision – BWG) Countercyclical capital buffer (CCB) of 0 Countercyclical capital buffer Systemic risk buffer (SRB) of up to 2 Other systemically important institutions buffer (O-SIIB) of up to 2 The table does not contain any credit control agreement or related amendments but only identifies milestones. “[year] to major amendment” indicates that the provision was or will be in place until the next major amendmentchanges listed may here have (minor been implemented in the meantime). 1977/11 to major 1977/11 to amendment suspension 1980/03 to 1981/03 1973/06 to major 1973/06 to amendment Limes (asset-side quantitative credit controls) Limes (asset-side quantitative credit Ultimate objective Intermediate objectiveIntermediate Legal basis Duration 1982/04 Gentlemen’s Agreement Gentlemen’s objective Ultimate objective Intermediate basis Legal Duration 1971/05 to 1980 currency lending foreign curb Measures to objective Ultimate objective Intermediate basis Legal Duration 2003 to 2013 2010 2013 to present2013 2006, 2011 Sustainability package objective Ultimate objective Intermediate basis Legal Duration 2013/01 to2013/01 present Macroprudential capital buffers Macroprudential objective Ultimate objective Intermediate basis Legal Duration 2016/01 to present 2016/01 to present 2016/06 to present Annex: Milestones of macroprudential policy in Austria since the 1950s and Chronik der Währungspolitik. Source: OeNB (1951–1983) Note:

MONETARY POLICY & THE ECONOMY Q3– Q4/16 189 The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator1

Stefan W. Schmitz1 In response to the end of the Bretton Woods system, the Oesterreichische Nationalbank (OeNB) aimed at keeping the value of the Austrian schilling stable vis-à-vis the currencies of the country’s main trading partners. The OeNB’s solution for achieving this objective was to introduce an internal target for the exchange rate between the Austrian schilling and the U.S. dollar (USD) based on a method called the Indicator. This paper investigates the origins of this Indicator. It finds that it was an ad hoc innovation developed by the OeNB during the days before August 25, 1971, in response to political opposition to its initial plans to revalue the Austrian schilling in line with the Deutsche mark and the Swiss franc only. The Indicator enabled the central bank to channel the political opposition into a technical debate and to make some concessions on details such as the composition of the currency basket. At the same time, it helped manufacture consent regarding the cornerstones of the new strategy, namely the stability of exchange rates vis-à-vis the main trading partners and a managed float vis-à- vis the U.S. dollar. The OeNB thereby preserved the traditional consensus in Austrian exchange rate policy, despite the fundamental shift from the traditional export-oriented to a stability- oriented exchange rate policy which took place in the early 1970s. JEL classification: E40, N10 Keywords: Austrian economic history, monetary and exchange rate policy, Bretton Woods, currency baskets The Oesterreichische Nationalbank Indicator were a success. Nevertheless, (OeNB) announced its new monetary very little is known about the Indica- policy strategy on August 24, 1971, one tor’s origin and its contribution to the week after the U.S.A. had closed the effectiveness of the OeNB’s new mone- gold window on August 15, 1971, and tary policy strategy. This paper aims to thus effectively put an end to the Bret- fill this void. ton Woods system. The OeNB’s new First, this paper traces the evolu- monetary policy strategy aimed at tion of the Indicator in August 1971. keeping the Austrian schilling (ATS) The analysis also takes into account the stable vis-à-vis the currencies of the politico-economic tensions among the country’s main trading partners. The stakeholders of Austrian exchange rate OeNB implemented this key objective policy. Furthermore, the study places by introducing an internal target for the Indicator in the wider context of the ATS/USD exchange rate based on a the transition from an export-oriented currency basket called the Indicator. to a stability-oriented exchange rate The literature concludes that the policy2, which took place in Austria in new monetary policy strategy and the the early 1970s. The investigation

1 Oesterreichische Nationalbank, Financial Stability and Macroprudential Supervision Division, stefan.schmitz@ oenb.at. The views expressed in this paper are exclusively those of the author and do not necessarily reflect those of the OeNB or the Eurosystem. The author would like to thank the referee, Eduard Hochreiter, Walpurga Köhler-Töglhofer, Helmut Pech, Karl Socher, Beat Weber and the participants in a workshop among the authors of this volume for their very useful comments as well as Walter Antonowicz (OeNB archive), the staff of the OeNB library and of the archives of the various other institutions mentioned in the text for their invaluable support throughout the research project. The author would also like to thank Emilia Haussteiner, Graeme High, Rita Schwarz and Susanne Steinacher for language advice and editing support. All remaining errors are the author’s. Refereed by: 2 A stability-oriented exchange rate policy is defined as one that puts higher weight on preventing the import of Tobias Straumann, inflation rather than on subsidising exporters and tourism. Until the early 1970s, the latter objective dominated University of Zurich over the former.

190 OESTERREICHISCHE NATIONALBANK The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

builds on expert interviews, material dize exports and tourism.5 Both politi- from the archives of the Republic of cal parties and the social partners were Austria, the OeNB, the social partners, committed to fixed exchange rates and the Social Democratic Party of Austria low inflation (Socher, 1973). All of (SPÖ) and the conservative Austrian these players were represented on the People’s Party (ÖVP). General Council of the OeNB. The The paper is structured along the Nationalbank Act of 1955 (National- following lines: Section 1 sketches the bankgesetz – NBG 1955) defined the economic and political environment in objectives of the OeNB in Article 2 (3) 1971. Section 2 describes the OeNB’s as follows: “It shall ensure with all the reaction to the end of Bretton Woods means at its disposal that the value of and the introduction of the Indicator. the Austrian currency is maintained Section 3 examines two hypotheses with regard both to its domestic pur- regarding the origin of the Indicator. chasing power and to its relationship Building on this, section 4 provides a with stable foreign currencies.” In detailed chronology of the development 1971, the SPÖ formed the Austrian of the Indicator prior to August 25, government; it had close ties with the 1971. Section 5 offers an assessment of Chamber of Labour and the Austrian the OeNB’s reaction to the end of Trade Union Federation. The ÖVP was Bretton Woods. Section 6 explains how the main opposition party, with close the Indicator contributed to the success ties to the Chamber of Agriculture, the of the OeNB’s new monetary policy Chamber of Commerce and the Feder- strategy. Finally, section 7 concludes. ation of Austrian Industries (FAI). Broad political backing and its 1 The economic and political clearly defined objectives underpinned environment in Austria in 1971 the OeNB’s credibility in Austria and A broad political consensus among the spurred the subordination of other two relevant political parties3 and the policy areas, i.e. fiscal, income, macro- social partners4 underpinned monetary prudential6 and capital account policy, policy and exchange rate policy in to the objectives of fixed exchange rates Austria in the post-war period and low inflation (Korp, 1971). (Mooslechner et al., 2007). Austria’s Increasing tensions within the successful long-term development Bretton Woods system characterized strategy after 1945 comprised the the international economic environ- internationalization of the Austrian ment. Growing international demand economy. In 1953, the Austrian for U.S. dollar liquidity led to a U.S. schilling became convertible: the ex- balance of payments deficit and exchange rate was fixed at ATS/USD 26, posed the U.S.A. to a run on its gold which implied a strategic undervalu- reserves (Triffin dilemma).7 The U.S. ation of the Austrian schilling to subsi- Administration took a number of mea-

3 Between 1945 and 1986, the SPÖ and the ÖVP regularly won more than 80% of the votes in general elections. 4 The social partners consist of the Chamber of Agriculture, the Chamber of Commerce, the Chamber of Labour, the Federation of Austrian Industries and the Austrian Trade Union Federation. 5 See Handler’s contribution to this volume for a history of Austrian exchange rate policy. 6 For details on the changing history of macroprudential policy in Austria after 1945, see Döme et al. in this volume. 7 By 1963 the combined U.S. dollar reserves of non-U.S. monetary authorities exceeded U.S. gold reserves (Eichengreen, 2008).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 191 The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

sures to address the symptoms of the 1971 the wider-than-expected U.S. underlying structural problem, such as balance of payments deficit again capital controls and Roosa bonds (U.S. boosted speculative capital flows out of government bonds guaranteed against the U.S. dollar. In August 1971, the U.S. dollar devaluation). Both, exter- U.S. Administration temporarily sus- nal and internal devaluations aimed at pended the convertibility of the U.S. bringing about balance of payments dollar into gold and other reserve assets adjustments were politically costly to protect its gold reserves. This effec- (Garritsen de Vries, 1976a). In addi- tively ended the Bretton Woods sys- tion, the U.S. monetary and fiscal ex- tem, although the Smithsonian Agree- pansion associated with the Vietnam ment in December 1971 prolonged its War and President Johnson’s Great demise until 1973 (James, 1996). Society agenda put further pressure on Officially the OeNB did not actively the U.S. balance of payments and on engage in this international reform dis- U.S. inflation. James (1996) and cussion; but its President at the time, Eichengreen (2008) argue that addressing Wolfgang Schmitz, called for strength- the structural problem would have ening the multilateral structure of the required diminishing the U.S. dollar’s international exchange rate system as reserve currency status – something opposed to unilateral actions (dual ex- the U.S.A. refused to do.8 The coun- change rate systems9 and revaluations) tries that held increasing amounts of (see i.a. Schmitz, 1970, 1972a, 1973). U.S. dollar reserves feared that the en- He advocated more flexibility in ex- suing increase in the volume of their change rates as well as in the use of the own currency in circulation would lead following methods: crawling pegs to inflation. In the late 1960s, the com- would be suitable to offset different bination of the progressive liberaliza- rates of inflation, the widening of mar- tion of the short-term capital account, gins around U.S. dollar parity would the fast-growing European dollar mar- be more appropriate to discourage ket and the resulting increase in the speculative capital flows, while (man- volatility of speculative capital flows in aged) floating rates could avoid specu- Europe further exacerbated the conse- lation in the period between the insight quences for European currencies that a currency is under-/overvalued (Schmitz, 1969). Fixed exchange rates, and the fixing of a new parity vis-à-vis capital mobility and an independent gold and the U.S. dollar. monetary policy proved hard to recon- In 1971, the OeNB partly and tem- cile (James, 1996). In May 1971, a porarily reversed the ongoing liberal- number of European countries, includ- ization of international capital flows. In ing Austria, Germany and Switzerland, 1946 the Foreign Exchange Act (Fed- temporarily floated or revalued their eral Law Gazette No. 162/1946) had currencies. Solomon (1977) suggests introduced strict capital controls in that, nevertheless, in early summer Austria. For all transactions in foreign

8 Nevertheless, in 1967 the members of the IMF agreed to introduce an artificial international means of payment, the Special Drawing Rights (SDRs), a basket currency that the IMF should have been able to allocate. However, the IMF allocated the first SDRs only in 1970 and its share in international reserves remained low (Garritsen de Vries, 1976a). 9 A dual exchange rate system features different exchange rates and policy regimes for different currency transactions, e.g. a fixed exchange rate regime for trade-related transactions and a (managed) floating rate regime for currency transactions related to the capital account.

192 OESTERREICHISCHE NATIONALBANK The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

currency and gold, prior approval had by foreigners. These measures proved to be sought from the OeNB, which effective, and the applications for ap- acted as foreign exchange authority. proval of capital imports by banks The approval practice was very strict remained low after these measures had until 1953, when the OeNB started a been taken. The third and final liberal- cautious and stepwise liberalization ization phase commenced in 1981 and process (Mooslechner et al., 2007). In ended in 1991, when capital controls the first phase (1954–59), it liberalized were finally abolished. current account transactions with the The OeNB closely monitored other member countries of the Organization countries’ reactions to the end of for European Economic Co-operation Bretton Woods. On August 21, 1971, (OEEC), Canada and the U.S.A. In the the Austrian representative to the second phase (1959–63), the OeNB International Monetary Fund (IMF) granted general approval for capital circulated an overview of the immediate account transactions to foreigners and exchange rate measures taken by other eased the approval practice in place for European countries (OeNB, 1971b, Austrians. The OeNB implemented the No. 778, annex 1). France would intro- liberalization process by issuing official duce a dual exchange rate (with the announcements according to the IMF’s approval, Garritsen de Vries Foreign Exchange Act, without seeking 1976a, p. 543). On the official French amendments to the act itself. That franc market, foreign currencies would ensured the OeNB a high degree of be bought and sold at par, on the finan- flexibility in tightening capital controls cial French franc market, the exchange if and when necessary. This was the rate would fluctuate according to sup- case when it again required individual ply and demand. The former was re- approval for bank deposits by foreigners served for transactions in connection in May 1971 and for other forms of with trade in goods and services and capital imports in November 1972 transactions of governments and public (until 1976). However, vis-à-vis banks, authorities. The latter was reserved for the OeNB largely refrained from im- capital account transactions. Belgium posing quantitative limits on capital im- and Luxembourg maintained their dual ports or exports and from tightening its exchange rate system, but stopped approval practice significantly. Instead ensuring that the margins of fluctua- it aimed at driving a wedge between in- tions on the official market were ad- terest paid on deposits in Austria and hered to. However, the economic union abroad through a so-called Gentlemen’s between Belgium and Luxembourg Agreement (see Döme et al., in this attempted to limit the fluctuations of volume). Similarly, quantitative restric- their common currency, the Belgian tions on capital imports via purchases of franc, vis-à-vis the Dutch guilder to domestic bonds and equity by foreigners 1.5%. Further internal OeNB docu- from Austrian residents were imple- ments analyzed the measures taken by mented in the form of amendments to other countries (OeNB, 1961–1978): the (voluntary) credit control agree- some suspended the buying rate of their ments in October 1972 (see Döme et currencies versus the U.S. dollar (Den- al., in this volume). In November 1972, mark, Portugal), while others informed the OeNB re-established the practice of the IMF that their currencies might requiring individual approval for pur- fluctuate outside the margins of ±0.75% chases of real estate and loans provided around parity (Germany, Japan, Spain

MONETARY POLICY & THE ECONOMY Q3– Q4/16 193 The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

and Switzerland). The countries in the OeNB approval) (OeNB, 1971a, 13/71); pound sterling block (e.g. Australia and and (3) the discontinuation of the New Zealand) aimed at keeping fluctu- OeNB’s official daily exchange rate an- ations within bandwidths vis-à-vis the nouncements according to the Foreign pound sterling, but not vis-à-vis the Exchange Act (OeNB, 1971a, 14/71). U.S. dollar. No country aimed at main- The central bank also declared its in- taining the value of its currency stable tention to stabilize the relationship of relative to a currency basket as an the Austrian schilling with the curren- immediate response to the events of cies of the major European industrial August 15, 1971. countries. Finally, it confirmed its commitment to multilateralism and 2 The OeNB’s reaction to the end close cooperation with the IMF. Over- of Bretton Woods: managed all, the OeNB adopted a strategy of floating and the Indicator managed floating for the ATS/USD On August 15, 1971, U.S. President rate. It decided against a dual ATS/ Nixon temporarily suspended the con- USD rate, as this would have violated vertibility of the U.S. dollar into gold the Articles of Agreement of the IMF and other reserve assets.10 This viola- and would have implied high bureau- tion of the Articles of Agreement of the cratic costs due to extensive current IMF effectively put an end to the multi- account and capital account controls, lateral exchange rate system of Bretton which could have hampered interna- Woods. Governments around the world tional trade. responded by closing currency markets The OeNB cited as the motives for for the following week. The authorities its new strategy its legal mandate pur- had to come up with new exchange rate suant to Article 2 (3) NBG 1955 and policies and monetary policy strategies. the broad societal consensus regarding On August 23, 1971, the OeNB of- price stability.12 ficially announced its reaction to the The OeNB operationalized its end of Bretton Woods. It suspended the objective of stabilizing the Austrian official buying rate of ATS/USD 24.51, schilling against the currencies of the but maintained the official selling rate major European industrial countries of ATS/USD 24.99.11 The gold parity and adopted the Indicator, a currency remained unchanged. In addition, the basket consisting of the currencies of its OeNB communicated a set of accom major trading partners. It did not com- panying measures: (1) the sterilization municate externally the currencies in of speculative capital inflows after the basket, nor their weights, nor the August 13, 1971; (2) the temporary daily target values.13 tightening of capital controls (capital The managed float was in operation account transactions with foreigners until the realignment of international were again made subject to individual exchange rates on December 21, 1971,

10 For accounts of the international negotiations before August 1971, see Solomon (1977, chapter 10) and Garritsen de Vries (1976a, chapter 24). 11 OeNB (1971d, August 23) and OeNB (1971e, August 23). 12 OeNB (1971d, August 23) and OeNB (1971e, August 23). 13 For two years the details of the method were kept confidential; Socher (1973) was the first to publish some of the details.

194 OESTERREICHISCHE NATIONALBANK The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

through the Smithsonian Agreement14. OeNB fixed the ATS/USD rate at On December 22, 1971, the OeNB 24.52. This served as the initial value announced a new ATS/USD central for the ATS/USD index. On the morn- rate of 23.30 and a broader bandwidth ing of August 25, 1971, the OeNB (±2.25%), but said it would aim for observed the following decline in USD ±1.5% under normal conditions (APA, rates for the basket currencies: –DEM 1971, December 22). The gold parity 0.15%, –CHF 0.25%, –ITL 0.41%, – remained unchanged (OeNB, 1971b, GBP 0.29%, –NLG 0.13% and SEK No. 795). Relative to the long-term 0%. The weighted sum (according to parity of ATS/USD 26, this constituted the above currency weights) amounted an appreciation of 11.59%, and a rise of to –0.20%. Thus, the so-called conver- 6.22% relative to the appreciation of sion factor was 99.80% (100%– May 9, 1971. The OeNB derived the 0.20%). The target rate was calculated new central rate from the Indicator as the product of the initial index value value as of December 22, 1971 (OeNB, times the conversion factor: (ATS/ 1971e, and APA, 1971, December 22). USD 24.52×99.80%=ATS/USD The Indicator was defined as the 24.47). Based thereon, the OeNB volume-weighted average of daily changes defined the pre-open trading band- of the exchange rates of major Euro- width as ATS/USD 24.44 to ATS/USD pean trading partners against the U.S. 24.54. The mid-rate was set at the mid- dollar, expressed in ATS/USD terms day fixing within the bandwidth based (OeNB, 1971b, No. 781). The list of on demand and supply on the Vienna main trading partners comprised Bel- foreign currency market. The OeNB gium, Denmark, the Federal Republic profited from the earlier daily price of Germany, Italy, Luxembourg, the discovery on the currency markets of Netherlands, Norway, Sweden, Swit- the currencies in the basket. zerland and the U.K. Trade with these nine countries amounted to 66.6% of 3 The origin of the Indicator Austria’s exports and imports of goods For the first time, the President of the (average for the period from 1968 to OeNB unofficially communicated the 1970). The trade shares were: Federal new monetary policy strategy on Au- Republic of Germany 33.4%, Switzer- gust 22, 1971, only one week after the land 8.5%, Italy 8.2%, the U.K. 6.3%, suspension of USD convertibility. This the Netherlands 4.6% (including Bel- section traces the roots of one of the gium and Luxembourg) and Sweden strategy’s major innovations, the cur- (including Denmark and Norway) rency basket called the Indicator. It 5.6%. Based thereon, the shares of the focuses on two hypotheses: respective six currencies in the Indica- 1. The Indicator might have been used tor amounted to: DEM 50.2%, CHF internally before August 1971, when 12.8%, ITL 12.3%, GBP 9.5%, NLG the costs and benefits of potential 6.9% (for Belgium, Luxembourg and realignments of the ATS/USD rate the Netherlands) and SEK 8.3% (for had been discussed internally. Denmark, Norway and Sweden). 2. The Indicator could have been an ad The calculation of the first Indica- hoc policy innovation that was tor value for August 25, 1971, can serve developed in the few days between as an example: on August 24, 1971, the August 15 and August 22, 1971.

14 For further details, see Solomon (1977, chapter 12), Garritsen de Vries (1976a, chapter 27) and James (1996, p. 235).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 195 The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

At first sight, hypothesis 1 seems more flation.15 The representatives of export- plausible. A new monetary policy strat- ers (Chamber of Commerce and FAI) egy implies substantial risk for policy- as well as the labor representatives makers (Schubert and Theurl, 1995), (Chamber of Labour and Austrian which incentivizes them to rely on Trade Union Federation) opposed the tested instruments. This hypothesis can appreciation.16 Eventually, consensus be checked by studying the internal emerged among all relevant players to documents concerning the OeNB’s refrain from ATS appreciation (OeNB, reactions to the appreciations of the 1969–1973, No. 258). Deutsche mark (DEM) in October Only one year later, in August 1969 and May 1971, as well as to the 1970, the OeNB reached a completely potential appreciations of the Deutsche different conclusion under the same mark in November 1968 and August scenario; rumors suggested that the 1970. Deutsche mark would appreciate again, The Austrian schilling did not and the OeNB prepared for a response. appreciate against gold and the U.S. The OeNB estimated the costs of dollar when the Deutsche mark unilat- imported inflation to be of a similar erally appreciated in November 1968 magnitude as in 1968 and 1969. But and October 1969. Internal OeNB this time, it regarded further subsidies analyses studied the effects of a DEM/ of exports and tourism unnecessary USD appreciation under the assump- and even detrimental to the objective of tion that the ATS/USD rate would not export market diversification. In addi- appreciate. They focused on the effects tion, investment activity was high at (1) on trade in goods and on tourism, the time, partly funded by Deutsche (2) on imported inflation, and (3) on mark debt. The OeNB feared that an net external debt service costs (OeNB, increase of the ATS value of the DEM 1961–1978). The OeNB highlighted debt burden would have damaging the benefits of the nonappreciation of effects on investment and growth. A the Austrian schilling in the form of stability-oriented exchange rate policy subsidies for exports and tourism, in began to replace the export-oriented line with the long-term development exchange rate policy that had been in strategy and the broad political place since 1945. For the first time, the consensus in Austria. At the same time, broad political consensus regarding it estimated the associated costs in the Austrian exchange rate policy was form of imported inflation to be com- called into question. At the end of the paratively low (up to 60 basis points if day, the Deutsche mark did not appre- the DEM/USD rate increased by 5%), ciate and the OeNB’s analysis remained because it predicted substitution without political consequences. effects, discounts of German exporters In May 1971, the OeNB decided to and the effects of Austrian disinflationary revalue the Austrian schilling against measures to cushion the impact of the U.S. dollar and gold by 5.05% – DEM appreciation. In 1969, the Aus- roughly in line with the Deutsche mark trian government initially favored an and the Swiss franc – and to widen the ATS appreciation as a means to fight in- bandwidth around U.S. dollar parity

15 Interestingly, the public blamed the government for high inflation, rather than the central bank (Socher, 1980). 16 Socher (1973) argues that the Austrian schilling would have appreciated with the Deutsche mark, if the Swiss franc had done so, too. Foreign policy concerns impeded an appreciation with the Deutsche mark only.

196 OESTERREICHISCHE NATIONALBANK The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

from 0.75% to 1%. The decision was which was so fierce that it led to con- based on the same reasoning as pre- flicts within the FAI Board (FAI, sented in August 1970. In addition, it 1971a, June 14). The critics complained was intended to fend off speculative heavily that the FAI had not done capital inflows and stem increasing enough to communicate the “very high pressure on wages from labor emigra- costs for exporters” of the ATS appreci- tion toward Switzerland and West ation to the government, the central Germany17 (SMP, 1971a). Responding bank and the general public. The FAI to demands by the FAI presented at the representative to the General Council General Council of the OeNB, the of the OeNB explained to his FAI President of the OeNB stated that bur- colleagues that speculative inflows dens for exporters and benefits for im- made appreciation inevitable. FAI porters were the inevitable conse- President Franz Josef Mayer-Gunthof quences of exchange rate revaluations; defended his administration against the compensation was not possible and harsh accusations and even cited from could not be expected from the central the minutes of the respective meeting bank, OeNB (1971f). of the OeNB General Council in May Representatives of the export in- 1971 to appease internal critics. In dustry – the Chamber of Commerce response, the FAI Board instituted a and the FAI – strongly opposed the Committee for Exchange Rate Policy, move (FAI, 1971a, June 29) and called which held its first meeting on Septem- for compensation (Die Industrie, 1971). ber 24, 1971 (FAI, 1971a, October 21). Nevertheless, the final decision was Neither internal OeNB documents backed by the government, the opposi- nor any of the minutes of OeNB meet- tion and the social partners (OeNB, ings contain any hint of the Indicator, 1969–1973, May 10, 1971); especially effective nominal exchange rates or the Austrian Trade Union Federation exchange rate indices. Thus, the pres- and the Chamber of Labour supported ent analysis rejects the hypothesis that the appreciation, because the Austrian the Indicator might have been used labor representatives traditionally had a internally at the OeNB before August low tolerance for inflation.18 Forming a 1971. consensus was helped by the fact However, the available documents that the effective ATS appreciation reveal increasing tensions regarding ex- amounted to only 3.3% rather than the change rate policy among the major official 5.05% and by government mea- stakeholders. In fact, the present analy- sures that eased the burden on exporters sis documents the transition from an (e.g. export subsidies, export guaran- export-oriented to a stability-oriented tees, and hedging operations at lower exchange rate policy in the early 1970s, rates). The OeNB’s Subcommittee on which unsettled the broad consensus Monetary Policy decided against float- that had governed Austrian exchange ing rates for fear of increasing specula- rate policy until then. tive capital flows. Testing hypothesis 2 – that the Indi- The final consensual decision did cator could have been a policy innova- not silence opposition among exporters, tion that was developed within a few

17 Higher wages in Switzerland and Germany led to labor shortages in western Austria and to undesired upward pressure on wages. 18 According to Schmitz (2016c), this was due to the interwar experience in Austria.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 197 The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

days in August 1971 – involved study- daily fixing at the Vienna stock ex- ing the minutes of internal OeNB change at 1:00 p.m. (Schmitz, 2016a). meetings and documents as well as the The few available documents do not minutes of the meeting of the Austrian provide details of how the new method Cabinet on August 23, 1971 (Austrian was derived by OeNB staff. However, Cabinet, 1971), and conducting expert the comparison of the Indicator weights interviews with contemporary witnesses. with the statistical material published The OeNB’s deliberations regarding in the monthly OeNB publication Mit- earlier events of actual and potential teilungen der Oesterreichischen Na- revaluations of the Deutsche mark are tionalbank (OeNB, 1971c; Communi- much better documented than the cations of the OeNB) sheds light on the events around the end of Bretton derivation of the details of the method, Woods. The minutes of the OeNB i.e. countries, weights and currencies Governing Board meeting on August involved. 22, 1971, state that “[b]ased on a novel First, table 4.1020 (OeNB, 1971c, method of calculation currently under February, p. 146–147 data on imports elaboration, the rates of the currencies and exports of goods) seems to have of the main European trading partners been the primary data source. Based should be kept stable…”19. On August thereon, the OeNB determined the 23, 1971, the OeNB presented to the weights by averaging each country’s government a proposal to revalue the share in total imports and exports of Austrian schilling by an amount to be goods from 1968 to 1970. These fig- derived from the Deutsche mark and ures excluded trade in services (tour- the Swiss franc appreciations, once the ism) although the sum of imports and currency markets had reopened (Aus- exports of services equated to 21% of trian Cabinet, 1971; see below for more the sum of import and exports of details). On August 25, 1971, the term goods. Less detailed statistics regarding Indicator appeared for the first time in the origin and destination of tourists the minutes of a Governing Board were available, however (table 4.2, meeting (OeNB, 1971b, No. 781). Two OeNB, 1971c, February, p. 145; data OeNB experts (Ferdinand Hain and Jo- on tourism, incoming: only guest hann Stelzer, Head and Deputy Head of nights (not expenditure) per country of the International Department, respec- origin; outgoing: only guest nights in tively) presented the general method aggregate, not according to destina- and weights to be used and gave exam- tion). Furthermore, the inclusion of ples. The Banking Department (Head: tourism data would have further in- Klaus Mündl) would calculate the daily creased the weight of the Deutsche Indicator values and fix the daily pre- mark in the basket, as the share of West open bandwidth with the Chief Execu- Germans among tourists visiting tive Director. In parallel, the Arbitrage Austria was very high from 1968 to Office would collect the bid and ask 1970 (75% of guest nights). The next prices and volumes from the Austrian section will highlight why this was a con- banks until midday, before joining the vincing argument for the OeNB.

19 „Mittels einer in Ausarbeitung befindlichen neuen Berechnungsmethode sollen in Hinkunft die Devisenkurse für die Währungen der wichtigsten europäischen Handelspartner möglichst stabil gehalten werden.“ (OeNB, 1971b, No. 778, August 22; author’s translation in the running text). 20 The data were collected by Statistics Austria from customs offices (country of origin) on a monthly basis rather than by the country of the trade partner (balance of payments).

198 OESTERREICHISCHE NATIONALBANK The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

Second, the countries considered in bilateral clearing, often in U.S. dollars. the index consisted of all member states The U.S.A. was excluded because the of the European Economic Community U.S. dollar was the index base, and its (EEC) except France, and all member inclusion would have had no effect states of the European Free Trade Area other than reducing the weight of the (EFTA) except Portugal. The primary currencies of the main trading partners data source (table 4.10 mentioned which the OeNB wanted to focus on. A above) actually contained data for all high weight of the U.S. dollar would EEC and EFTA countries plus the have undermined the objectives of ex- U.S.A., the countries participating in change rate realignment: stemming the Council for Mutual Economic speculative capital inflows, preventing Assistance (Comecon) and Yugoslavia. a wage drift due to labor emigration to France was excluded because of the West Germany and Switzerland and dual exchange rate it introduced in fighting the import of inflation via the August 1971, according to Hochreiter inflow of U.S. dollars and higher prices (1975, p. 10), and/or because of its of imports from the main trading part- small share of total trade in goods, ac- ners. The rest of the world accounted cording to Schmitz (1972b, p. 151). for the remaining 10% of Austrian The group around the Dutch guilder trade in goods. (Belgium, Luxembourg and the Neth- Third, the currencies included are erlands) constituted the smallest weight those of the countries in the basket (4.6%) among the countries considered except for Belgium, Luxembourg, in the Indicator. France would have Norway and Denmark. Luxembourg commanded a share of only 3.4% of shared a currency with Belgium Austrian trade (author’s calculations (OeNB, 1971b, No. 778, annex 1).21 based on table 4.10). Portugal was not Norway and Denmark were grouped considered because of the small size of together with Sweden in the Swedish bilateral trade (0.7% of total trade in crown (SEK) group (5.6%). The avail- goods; author’s calculations). The coun- able documents do not provide the rea- tries considered in the Indicator ac- soning behind this decision, but pre- counted for 66.6% of Austrian trade in sumably their individual trade shares goods. Even if France and Portugal had were too small (Norway below 1%; been considered, 30% of Austrian trade Denmark 1.7%; author’s calculations). in goods at the time would not have been covered by the Indicator. Of this 4 The politico-economic tensions share, Eastern European Comecon created by Austria’s reaction to member states accounted for about the end of Bretton Woods and 11 percentage points, Yugoslavia for the emergence of the Indicator about 3 percentage points and the The following section examines the ar- U.S.A. for about 4 percentage points. gument that the OeNB developed the The Comecon countries and Yugoslavia Indicator in response to opposition to were excluded from the Indicator be- its initial reaction to the end of Bretton cause their currencies were not con- Woods. What were the policy stances vertible and trade was largely based on of the major stakeholders?

21 In addition, the Belgian franc was in any case pegged to the Dutch guilder.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 199 The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

4.1 Policy stances of major stake- on August 5, 1971, because of foreign holders in Austrian exchange policy concerns. Also the representa- rate policy tives of credit institutions in the Gen- The broad consensus that had governed eral Council of the OeNB expressed a Austrian exchange rate policy after 1945 strong desire to avoid the impression of was seriously put to the test in August strong dependence on German eco- 1971. The policy stances of the major nomic and monetary policy.23 stakeholders were hard to reconcile: On August 18, 1971, Austrian • The Ministry of Finance and the rep- newspapers (Kurier, 1971a; Die Presse, resentatives of exporters regarded 1971) reported that the Minister of the ATS appreciation of May 1971 as Finance wanted to introduce a dual ex- sufficient. They wanted to keep the change rate regime. Following the ex- ATS/USD exchange rate within the amples of France and Belgium, Austria bandwidth of ATS/USD 24.51 to would distinguish between an ATS/ ATS/USD 24.99. The OeNB argued USD exchange rate for trading in goods for a greater flexibility of exchange and services and an ATS/USD ex- rates (OeNB, 1971b, No. 778). change rate for financial transactions. • The Minister of Finance, Hannes The OeNB opposed such a move as it Androsch, favored a dual exchange would constitute a step backward in rate system (like the ones introduced the process of multilateralism and fi- in France and Belgium) – an approach nancial integration. It would also con- which was rejected by the OeNB and flict with the Articles of Agreement of the main opposition party.22 the IMF (although the IMF had ap- • The Minister of Finance opposed proved the dual exchange rate system managed floating, while the OeNB was for France). The substantial administra- in favor (Kurier, 1971b, August 23). tive burden would impede international payments (SMP, 1971b; OeNB, 1971b, 4.2 Chronology of the political No. 778). deliberations before the On August 22, 1971, the Minister reopening of the foreign of Finance and the OeNB President currency market in Vienna met to find a consensus. No documen- Some of the stakeholders of Austrian tation of the meeting could be retrieved exchange rate policy proposed alterna- from the Austrian State Archives or the tives to managed floating based on the OeNB’s archives. The position of the Indicator. This section provides a OeNB participants in the meeting had chronology of the political delibera- been agreed in a Governing Board tions up to August 25, 1971, aiming to meeting on the same day (OeNB, 1971b, reconstruct the individual steps that led No. 778): advocating the reopening of to the OeNB’s response to the end of the foreign currency market in Vienna Bretton Woods. on August 24, 1971, a managed float The option of a simple DEM peg against the U.S. dollar and stable ex- had already been rejected by the OeNB change rates vis-à-vis the main trading

22 Kurier (1971a, August 25); Die Presse (1971, August 18); Arbeiterzeitung (1971, August 19); Wiener Zeitung (1971, August 24); Volkszeitung Kärnten (1971, August 25); Börse-Zeitung (1971, August 25). 23 See SMP (1971b), Schmitz (2016b and 2016c). In the latter, Heinz Kienzl reported that the U.S.S.R. trade envoy in Vienna had voiced concerns regarding a peg of the Austrian schilling to the Deutsche mark and feared it would constitute a step toward the reunification of Austria and Germany.

200 OESTERREICHISCHE NATIONALBANK The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

partners. In its August 23, 1971, edi- As a consequence, on August 24, tion, the Austrian daily Kurier (1971b) 1971, the OeNB added the Italian lira reported on the meeting, suggesting (ITL) and the pound sterling (GBP) to that a consensus had been found and the Deutsche mark and the Swiss franc quoting the OeNB President: “The for calibrating the target value for the OeNB’s efforts are directed toward the ATS/USD exchange rate (indirect evi- smallest possible changes to exchange dence in OeNB, 1971b, No. 781). The rates vis-à-vis trading partners.”24 likely reason for the broadening of the On August 23, 1971, the Minister basket of currencies was that exporters of Finance presented the OeNB pro- feared that the Deutsche mark and the posal in an extraordinary Cabinet Swiss franc would appreciate further, meeting (Austrian Cabinet, 1971) in which would have a negative impact on the presence of the OeNB’s executive exports and tourism. comittee. Like in May 1971, the ATS On August 25, 1971, the term Indi- exchange rates vis-à-vis the main trad- cator appeared for the first time in an ing partners were to be stabilized by fo- internal OeNB document (OeNB, cusing on relatively stable ATS/CHF 1971b, No. 781): “From now on, the and ATS/DEM rates. The Minister of calculation of the so-called Indicator Finance stated that “[t]his would be covers about two-thirds of trade practically identical to an average rate through the inclusion of the Benelux across European currencies.”25 He also countries and three Scandinavian coun- announced that the OeNB would fix tries.”26 This wording suggests that the ATS/USD rate on August 24, 1971, coverage of the narrower currency within the bandwidth of ATS/USD group (DEM/CHF/ITL/GBP) was still 24.51 and ATS/USD 24.99 to ease the considered to be too low. The Deutsche burden on exporters and the tourism mark and the Swiss franc had appreci- sector. ated more than other currencies over Immediately after this meeting, the the previous years, so that the reduc- Federal Chancellor, the Minister of tion of their weights would address the Finance and the OeNB’s executive concerns of the FAI and of the Chamber comittee met with the social partners of Commerce. This also clarifies why (Austrian Cabinet, 1971). No docu- the OeNB considered the high weight mentation of this meeting could be re- of Germany in the share of incoming trieved from the archives of any of the tourism (75%) a convincing reason for participating institutions. However, in- basing weights calculation on trade in direct evidence suggests that the com- goods only, as discussed above. It would mon proposal by the government and have strongly increased the weight of the OeNB was subject to fierce opposi- Germany even in a broader basket. In tion from the FAI and the Chamber of addition, the OeNB fixed the ATS/ Commerce. USD rate at 24.52 “[t]o fulfill the com-

24 „Das Bemühen der Notenbank gehe dahin, sagte Dr. Schmitz, gegenüber den Handelspartnern möglichst geringe Änderungen der Wechselkurse zu erreichen.“ (Kurier, August 23, 1971b; author’s translation in the running text). 25 „Das kommt praktisch einem Mischkurs aus den Währungen in Europa gleich.“ (Austrian Cabinet, 1971, p. 2; author’s translation in the running text). 26 „Bei der Berechnung des sogenannten Indikators sind nunmehr durch die Einbeziehung der Benelux-Staaten und drei der skandinavischen Länder rund zwei Drittel der Außenhandelspartner (sic) berücksichtigt.“ (OeNB, 1971b, No. 781, p. 2; author’s translation in the running text).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 201 The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

mitment [of August 23] to the govern- as small as possible vis-à-vis trading ment and the social partners that the partners. The OeNB framed the Indi- Austrian export industry would not be cator as the “operationalization” of an burdened further, as far as possible.”27 exchange rate policy it had decided on Based on the OeNB internal estimates and communicated before the relevant of the ATS/USD rate of 24.38 as of method had been developed. August 23, 1971 (OeNB, 1971b, No. Little detailed internal OeNB analysis 779), the fixing of the initial index is available regarding the development of value at ATS/USD 24.52 reduced the the Indicator. Nonetheless, it is likely that long-term appreciation effect by about the introduction of a currency basket was 0.6% – a sizeable easing of the burdens inspired by an IMF staff paper on ex- on exporters. In addition, the OeNB change rate indices (Hirsch and Higgins, reduced the costs of short-term hedg- 1970).28 Ferdinand Hain, who presented ing (via changes to the Gentlemen’s the Indicator at the OeNB Governing Agreement), offered central clearing Board meeting on August 25, 1971, rou- for long-term hedging via government tinely held close contact with the IMF, es- U.S. dollar debt and took on the for- pecially with Rudolf Rhomberg, a fellow eign currency risk from exporters Austrian, who served as head of the IMF through its daily market interventions Research Department. Fred Hirsch, also (Schmitz, 1972b). an Austrian, was a Senior Adviser there.29 The chronology of the political deliberations shows that the introduc- 4.3 The Indicator after December tion of a currency basket of more than 1971 two currencies, its final composition of The daily ATS/USD exchange rates30 ten countries and six currencies, and on the Viennese foreign currency the initial index value emerged in market hardly deviated from the daily response to criticism from representa- Indicator value (Schmitz, 2016a). tives of the export industry and the The Indicator composition and tourism sector of the OeNB’s initial weights remained unchanged until June proposal to revalue the ATS/USD rate 22, 1972, when the pound sterling based purely on the development of the floated. The OeNB refined its interpre- DEM/USD and the CHF/USD rates. tation of its legal mandate to avoid a The OeNB adapted the method of im- conflict of interest between price sta- plementation, but not the initial strat- bility and exchange rate stability vis-à- egy it had communicated to the public vis countries with higher inflation for the first time on August 22, 1971, (namely to maintain the stability of the namely to keep exchange rate changes domestic purchasing power of the cur-

27 „[Obwohl die heutige vorbörsliche Entwicklung des US$ auf den wichtigsten europäischen Devisenmärkten eine deutliche Abwärtsentwicklung aufweist und die nicht mehr in Geltung befindlichen unteren Interventionspunkte fast überall unterschritten wurden, soll der $-Kurs auf der Wiener Devisenbörse bei Eröffnung noch oberhalb des unteren Interventionspunktes liegen,] um der Zusage gegenüber der Bundesregierung und den Sozialpartnern Rechnung zu tragen, daß der österreichischen Exportindustrie keinen zusätzlichen Bürden auferlegt werden.“ (OeNB, 1971b, No. 780, p. 1–2; author’s translation in the running text). 28 Schmitz (2016b and 2016d). 29 Hirsch and Higgens (1970) might have been prompted by the introduction of SDRs (see footnote 8) by the IMF in 1973, but no reference to SDRs is made therein. 30 For details regarding the development of the ATS/USD exchange rate and ATS/DEM exchange rate from 1952 to 1999 see chart 1 and chart 2 in Nowotny (2007).

202 OESTERREICHISCHE NATIONALBANK The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

rency and its exchange rate relative to toward the Deutsche mark, though no the stable foreign currencies; NBG formal ATS/DEM peg was announced. 1955, Article 2 (3)). It argued that a The OeNB kept the initial index currency that depreciated by floating value of ATS/USD 24.52 unchanged was not stable, and therefore excluded until March 16, 1973, when it rebased the pound sterling from the Indicator. the index value as the ATS/USD mar- Otherwise, the ATS/USD exchange ket value on February 9, 1973.32 The rate would have been pushed outside OeNB further rebased the initial index the bandwidth of 2.25% under the value (1) on July 3, 1973, when the Smithsonian Agreement. The decision Deutsche mark and the Austrian schil- was alleviated by the fact that exporters ling appreciated; (2) on May 17, 1974, bore few losses and some importers when the bandwidth of the Austrian profited (raw materials often traded in schilling increased to 4.5%, and (3) on pound sterling) (Socher, 1973). The July 10, 1975, when the French franc Italian lira was dropped from the basket returned to the “snake in the tunnel,” on March 15, 1973, when it floated. At i.e. the EEC exchange rate system in that time the Swiss franc was also ex- place from 1972 to 1979. cluded for a period of two weeks when The interpretation, as of June 22, it appreciated too much and would have 1972, of a “stable currency” as a cur- pushed the ATS/USD rate outside the rency that does not depreciate trig- bandwidth. To prevent the Deutsche gered an upward drift of the Deutsche mark weight from increasing too mark’s weight in the basket from 50.2% strongly, the OeNB added the French on August 25, 1971, to 63.4% on franc (FRF) to the basket after France January 22, 1974. In combination with had ended its dual exchange rate system the block-floating of the European cur- on March 16, 1973.31 Moreover, as of rencies in the basket, this led to that date the Danish crown (DKK) was Austria’s Deutsche mark orientation as included in the basket separately from of 1976 and to the emergence of the the Swedish crown (SEK). As of hard currency policy as of 1979 (Socher, September 17, 1973, the Belgian franc 1996, Mooslechner et al., 2007). (BEF) was separated in the basket from the Dutch guilder (NLG). 5 Assessment of the new As of March 1973, the Indicator monetary policy strategy consisted only of currencies that took of August 1971 part in the block floating around the The OeNB’s new monetary policy Deutsche mark, such that Austria effec- strategy of August 1971 – in concert tively – though not formally – joined with other economic policy initiatives that floating block. The dominant role – proved successful overall, though its of the Deutsche mark therein turned societal costs (e.g. in the form of subsi- the Indicator value of the ATS/USD dies for investment and exports) should exchange rate into a close shadow of not be ignored. the DEM/USD exchange rate. As of • The OeNB was able to maintain its July 13, 1976, the central rate for the exchange rate target. After the col- Austrian schilling was oriented solely lapse of Bretton Woods, many coun-

31 The French franc was dropped from the basket again when it was forced to float and devalue on January 22, 1974. 32 February 9, 1973 was the last trading day before the U.S. dollar devalued by 10% on February 10.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 203 The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

tries targeted fixed exchange rates. situation (Butschek, 1985). In re- However, most of them were forced sponse, Austria implemented nine to let their currencies float and even- stability packages between 1972 and tually devalue them; only three were 1977, which were coordinated across able to avoid involuntary devaluations policy areas, to address inflation and (Austria, Hong Kong and Saudi current account imbalances ex post. Arabia; Streissler, 1998). • The new monetary policy strategy • Austria’s macroeconomic perfor- also carried societal costs. Strau- mance after 1971 in terms of infla- mann (2010) suggests that the Aus- tion, per capita GDP growth, the rise trian response to relatively high real in the public debt-to-GDP ratio, and wage growth between 1973 and 1983 unemployment was very good by consisted not only of ex post stability international comparison (Butschek, packages. He argues that Austria also 1985; Hochreiter, 1981; Nowotny, reduced labor immigration, hoarded 2007). Labor productivity grew labor in state-owned companies and faster than in Austria’s major trading subsidized investments and exports partner countries. Handler (1989) (as compensation for exporters). This suggests that this was a result of the caused direct fiscal costs, capital mis- stability-oriented exchange rate policy allocation and an increase in nonper- due the competitive pressure it in- forming bank loans in the 1980s. duced on exporting sectors. • Austria did not experience a twin 6 Significance of the Indicator crisis when it liberalized its capital within the overall monetary account and its financial sector, policy strategy unlike many other countries The following assessment of the Indica- (Kaminsky and Reinhart, 1999) such tor takes the OeNB’s accompanying as Spain (1978), Denmark (1987), measures as given. The OeNB under- Norway (1989), Sweden and Finland pinned its exchange rate target with a (both 1991). Capital account liber complex set of instruments to safe- alization was subordinated to ex- guard technical implementation. These change rate policy; i.e. when neces- instruments included the flexible adap- sary the OeNB temporarily tightened tation of capital account liberalization, capital account controls. credit control agreements and the Gen- • Within Austria, the new monetary tlemen’s Agreement, the monopoliza- policy strategy was successful insofar tion of the ATS/DEM market, the pro- as it supported coordination across hibition for Austrian banks to short-sell different economic policy areas Austrian schillings, and high foreign (Glück et al., 1992; Haberler, 1979; currency reserves relative to foreign Guger, 1998). It did so mainly by holdings of Austrian schilling-denomi- subordinating other policy areas nated bonds (Streissler, 1998). (fiscal, income, monetary, capital Despite its technical shortcomings account and macroprudential policy) (Hochreiter, 1975), the Indicator was a under exchange rate policy (or the useful instrument for achieving the exchange rate target, respectively). objective of stabilizing the Austrian However, ex ante subordination was schilling’s exchange rate vis-à-vis the not perfect, and after 1973 wages in- currencies of its main trading partners creased faster than would have been in terms of nominal effective exchange compatible with the current account rates. The first internal OeNB assess-

204 OESTERREICHISCHE NATIONALBANK The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

ment of the Indicator on September 8, The Indicator contributed signifi- 1971, was positive and stated that the cantly to the success of the new mone- Indicator reduced the volatility of the tary policy strategy for three reasons: Austrian schilling against the major First, the OeNB overcame the trading partner currencies and in- opposition to its initial policy stance on creased volatility vis-à-vis the U.S. dollar August 23, 1971, by presenting the (OeNB, 1971b, No. 782). On Indicator as the operationalization of December 5, 1971, the Austrian gov- the two undisputed cornerstones of a ernment and the OeNB agreed that stability-oriented exchange rate policy: there was no need to change the policy (1) the OeNB’s legal mandate (NBG stance (OeNB, 1971b, No. 792; Meet- 1955, Article 2 (3), “It shall ensure ing of the Minister of Finance and the with all the means at its disposal that OeNB executive committee, Decem- the value of the Austrian currency is ber 6). The new parity agreed upon in maintained with regard both to its the exchange rate realignment on domestic purchasing power and to its December 22, 1971, was based on the relationship with stable foreign curren- Indicator, which an internal study cies.”) and (2) the interpretation of the found to be successful (OeNB Presi- currencies of Austria’s trading partners dent in APA, 1971, December 22). as (mostly) stable currencies. As a con- Over the medium term, it is interesting sequence, the key ingredient of the to note that despite a series of apprecia- OeNB’s reaction to Bretton Woods – tions of the Austrian schilling against “to keep the value of the Austrian the U.S. dollar, the Austrian schilling’s schilling stable vis-à-vis the main trad- nominal effective exchange rate (as ing partners” (OeNB, 1971d; OeNB, calculated in Hochreiter, 1981) only 1971e) – was largely undisputed as increased modestly until January 1973 well. The representatives of the export (by 1.2%). industry and the tourism sector could However, the strategy also gener- not reject these cornerstones, but ini- ated losses for the OeNB due to cur- tially strongly opposed the sole orienta- rency appreciation versus the U.S. tion of the Austrian schilling toward dollar, in which a large share of its re- the Deutsche mark and the Swiss franc. serves were denominated. Losses In response, the OeNB broadened the amounted to ATS 776 million (USD 33 basket of currencies and framed the million) from August 24 to November broader set as a sophisticated technical 30, 1971, and increased to ATS 2 bil- currency basket33 that simply opera- lion by December 13, 1971 (OeNB, tionalized (1) and (2) above. 1971b, No. 794). The OeNB planned Second, the Indicator forced the to cover the costs out of its revaluation political debate into a technical frame- reserves without a reduction of fiscal work. Its technical sophistication and seignorage (OeNB, 1971b, No. 792, scientific rigor34 lent credibility to that annex). framework (Schmitz, 2016d). Critics

33 A number of studies (inter alia Hochreiter, 1975) argue that the Indicator had technical shortcomings compared with more sophisticated nominal exchange rate indices. However, the OeNB did not aim to develop a highly sophisticated method. The level of sophistication just had to be high enough to fulfill the OeNB’s requirements in terms of its discursive strategy in the political debate. 34 The fact that it was inspired by academics and by a neutral international organization like the IMF could have lent additional weight to the Indicator. A similar strategy of international bond investors in London around the 1860s is described in Flandreau (2016).

MONETARY POLICY & THE ECONOMY Q3– Q4/16 205 The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

of the revaluation could not simply re- other end of the spectrum of the social ject the method as inappropriate or ar- partners, the Chamber of Labour em- bitrary. This discursive strategy in the braced the strategy, too: “It must be as- political debate proved successful. The serted that the exchange rate measures fiercest critics of the ATS revaluation of of the government and the OeNB turn May 1971, the FAI, embraced the new out to be appropriate.”36 The first publi- monetary policy strategy of August cation on the new strategy highlighted 1971. In its first meeting on September the politico-economic contribution 24, 1971, the FAI’s newly founded of (managed) floating and, indirectly, Committee on Monetary Policy for- of the Indicator: “The result of the mally stated: “In principle, the industry realignment of exchange rates [as of embraces the temporary concept of Au- December 22, 1971] were revaluations gust 20 (sic), 1971. It consists of the at- and devaluations with respect to the tempt to allow the Austrian schilling to situation before August 15 and even move with the major European curren- more so with respect to the situation cies (DEM, CHF, ITL, NLG, SEK (sic)) before May 10, [1971], that could never in their relationship to the U.S. dollar have been reached by way of negotia- and to embrace the possibility of a slip- tions.”37 The OeNB managed to over- ping of the U.S. dollar.”35 The Chamber come the politico-economic tensions of Commerce developed its own cur- between different stakeholders in Aus- rency basket to be able to challenge the trian exchange rate policy. Despite the Indicator in political deliberations conflicting views regarding revaluation (Socher, 1973). Its basket was much and nonrevaluation, managed float broader and included countries with versus fixed exchange rates or dual nonconvertible currencies; the weights exchange rates, the OeNB preserved of the currencies that were considered the broad consensus on exchange rate more likely to exert appreciation pres- policy. sure on the Austrian schilling (the Third, the Indicator contributed to Deutsche mark and the Swiss franc) the international credibility and the were much lower. As of July 1973, the broad societal backing of a stability-ori- Indicator appreciated by 12.4% vis-à- ented exchange rate policy, alongside vis the index of the Chamber of Com- the OeNB’s accompanying measures. merce (Socher, 1973, p.108). Never- The Indicator constituted a flexible and theless, the OeNB reached its objective nontransparent policy rule (neither the of moving the political debate onto a composition of the basket nor the technical level, while preserving the weights were publicly communicated). broad political consensus regarding the This seems to somewhat contradict the fundamentals of the strategy. At the conventional economic wisdom that

35 „Die Industrie steht grundsätzlich nach wie vor zu dem vorläufigen Konzept vom 20. (sic) August 1971. Dieses besteht in dem Bemühen, den österreichischen Schilling mit den wichtigsten europäischen Währungen (DM, sfr, Lire, hfl, SKr (sic)) in ihrem Verhältnis zum Dollar mitgehen zu lassen und ein Abgleiten des Dollar-Kurses in Kauf zu nehmen.“ (FAI, 1971a, October 21; author’s translation in the running text). 36 „Es muss festgestellt werden, dass sich die von der Bundesregierung und von der Nationalbank gesetzten Maßnahmen als richtig erwiesen haben.“ (Austrian Chamber of Labour, 1971; author’s translation in the running text). 37 „Das Ergebnis der Neufestsetzung der Wechselkurse waren gegenüber der Situation vor dem 15. August und noch mehr gegenüber der Situation vor dem 10. Mai Auf- und Abwertungen, die im Verhandlungswege nie erreicht werden hätten können.“ (Schmitz, 1972b, p. 290: author’s translation in the running text).

206 OESTERREICHISCHE NATIONALBANK The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

monetary policy rules need to be simple USD and the CHF/USD exchange and transparent to achieve their objec- rates. The OeNB adapted the method tive (Dorn, 1979; Hochreiter and of implementation but not its initial Winckler, 1995). But within the frame- strategy, which it had already commu- work of managed floating, this lack of nicated to the public for the first time transparency exposed potential specu- on August 23, 1971, before the Indica- lative capital flows to a two-way risk. tor was developed. The OeNB did not announce a target By implementing the Indicator, the value for the ATS/USD exchange rate OeNB forced the political discourse or the Austrian schilling’s exchange into a technical framework. This enabled rate vis-à-vis any other foreign cur- the OeNB to maintain its initial policy rency (ATS/DEM or ATS/CHF). As a stance of August 22, 1971, in the face consequence, there was no exchange of fierce opposition from exporters and rate target that could lend itself to a the tourism sector in view of the transi- speculative attack. This, in turn, tion from an export-oriented to a sta- reduced the costs of maintaining the bility-oriented exchange rate policy. target. Governments are forced to de- The OeNB managed to push through value the currency when the societal its policy stance despite the viewpoints costs of defending a target (that has lost of other stakeholders in Austrian ex- credibility) become unsustainably high change rate policy. It forged a consen- in terms of high interest rates, restric- sus (1) on managed floating (rather than tive fiscal policy, capital account con- fixed exchange rates or dual exchange trols and high unemployment. Austria rates) and (2) on the revaluation of the avoided such costs, which made it eas- Austrian schilling with respect to the ier to preserve the societal consensus U.S. dollar. In addition, the govern- and the political resolve. Over time, ment shouldered some of the costs of Austria’s stability orientation was this consensus in the form of compen- noticed internationally and this in- satory measures for exporters and, creased the credibility of the chosen more indirectly, through its general policy. Similarly, its economic benefits labor market policies, e.g. labor hoarding (i.e. exchange rate stability vis-à-vis the in state-owned enterprises. The con- major trading partners) materialized sensus supported policy coordination and reinforced the political resolve and across different economic policy areas, societal consensus within Austria. which in turn helped contain the Eventually, the ATS/DEM peg became societal costs such as high unemploy- a reality without the OeNB ever for- ment or currency crises. mally committing to a fixed exchange Finally, the Indicator contributed to rate vis-à-vis the Deutsche mark. the international credibility of a stability-oriented exchange rate policy, in 7 Conclusions combination with instruments that The present analysis shows that the deterred speculative attacks. Following Indicator was an ad hoc innovation in the Austrian schilling’s orientation to- August 1971. The OeNB developed it ward the Deutsche mark as of 1976, in response to the criticisms of repre- Austria’s hard currency policy emerged sentatives of the export industry and as of 1979 as a result of the OeNB’s the tourism sector to its initial proposal reaction to the end of Bretton Woods to revalue the ATS/USD exchange rate and the introduction of the Indicator. based on the development of the DEM/

MONETARY POLICY & THE ECONOMY Q3– Q4/16 207 The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

References Arbeiterzeitung. 1971. 23,50 Schilling für einen Dollar. Nationalbank sichert Zahlungsver- kehr – Brüssel neuer Brennpunkt. Thursday. August 19. 1. Austrian Cabinet. 1971. Verhandlungsschrift über die Sitzung des Ministerrats Nr. 64b (Minutes of the meeting of the Cabinet No. 64b). Austrian State Archives. August 23. Austrian Chamber of Labour. 1971. Protokoll über die am 30. September 1971 im Flughafen- restaurant Salzburg abgehaltene Vorstandssitzung des österreichischen Arbeiterkammertags. (Minutes of the Board Meeting of the Austrian Chamber of Labour on September 30, 1971, at the Salzburg airport restaurant). September 30. Archives of the Vienna Chamber of Labour. Austrian Press Agency (APA). 1971. Various articles. Börse-Zeitung. 1971. Österreich schlägt den Weg zur Mitte ein. Schilling soll vorerst nur gegen den Dollar schwanken. Wednesday. August 25. Butschek, F. 1985. Die österreichische Wirtschaft im 20. Jahrhundert. WIFO und G. Fischer Verlag. Vienna. Stuttgart. Csoklich, W., W. List and A. Schwarzer. 1987. Das österreichische Devisenrecht. Manz. Vienna. Denk, A. 1976. Wechselkurspolitik im Floating. Österreichisches Bank-Archiv 24/2. 53–65. Die Industrie. 1971. Nach der Schillingaufwertung. Die Industrie 71/20. May 14. 3–5. Die Presse. 1971. Doppelter Kurs für Dollar? Androsch: Keine Schillingaufwertung. Wien und Bern konform. Wednesday. August 18. 1. Döme, S., S. W. Schmitz, K. Steiner and E. Ubl. 2016. The changing role of macroprudential policy in Austria after World War II. In this volume. Dorn, H. 1979. Wechselkurspolitik ohne feste Spielregeln. Wirtschaftspolitische Blätter 6. 5–15. Eichengreen, B. 2008. Globalizing Capital. A History of the International Monetary System. Princeton University Press. Princeton. FAI. 1971a. Protokolle über die Sitzungen des Vorstands der Österreichischen Industriellen vereinigung (Minutes of the FAI board meetings). Archives of the Federation of Austrian Industries. FAI. 1971b. Protokolle über die Sitzungen des Präsidiums der Österreichischen Industriellen vereinigung (Minutes of the FAI executive committee). Archives of the Federation of Austrian Industries. Flandreau, M. 2016. Anthropology and finance: financial history of Victorian science. Keynote speech at the joint conference of the OeNB and the European Association for Banking and Financial History. Vienna. April 29. Garritsen de Vries, M. (ed.). 1976a. The International Monetary Fund 1966–71. The System Under Stress. Volume I: Narrative. IMF. Washington D.C. Garritsen de Vries, M. (ed.). 1976b. The International Monetary Fund 1966–71. The System Under Stress. Volume II: Documents. IMF. Washington D.C. Glück, H., D. Proske and J. A. Tatom. 1992. Monetary and Exchange Rate Policy in Austria: An Early Example of Policy Coordination. Federal Reserve Bank of St. Louis. WP 1992-005A. St. Louis. Guger, A. 1998. Economic Policy and Social Democracy: The Austrian Experience. In: Oxford Review of Economic Policy 14. 40–58. Haberler, G. 1979. Austria’s Economic Development after the Two World Wars: A Mirror Picture of the World Economy. In: Clement, W. and K. Socher (eds.). Empirische Wirtschafts- forschung und monetäre Ökonomik. Commemorative publication for Stephan Koren. Berlin. 176–199. Handler, H. 1983. Die österreichische Hartwährungspolitik. In: Abele, H., E. Nowotny, S. Schleicher and G. Winckler (eds.). Handbuch der Österreichischen Wirtschaftspolitik. Second edition. Manz. Vienna. 413–426.

208 OESTERREICHISCHE NATIONALBANK The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

Handler, H. 1989. Grundlagen der österreichischen Hartwährungspolitik. Manz. Vienna. Handler, H. 2016. Two centuries of currency policy in Austria. In this volume. Hirsch, F. and I. Higgins. 1970. An Indicator of Effective Exchange Rates. In: IMF Staff Papers. 453– 487. Hochreiter, E. 1975. Der österreichische Wechselkursindikator und das Konzept der effektiven Wechselkurse ad No. 354/75/1. OeNB Bank History Archives. Hochreiter, E. 1981. Wechselkurspolitik und Wettbewerbsfähigkeit. Eine längerfristige Analyse für Österreich 1967–78. Schriftenreihe der Bundeswirtschaftskammer 42. Vienna. Hochreiter, E. and G. Winckler. 1995. The advantages of tying Austria’s hands: The success of the hard currency strategy. In: European Journal of Political Economy 11/1. 83–111. James, H. 1996. International Monetary Cooperation since Bretton Woods. Oxford University Press. Oxford. Kaminsky, G. L. and C. M. Reinhart. 1999. The Twin Crises: The Causes of Banking and Balance-of-Payments Problems. In: American Economic Review 89/3. 473–500. Korp, A. 1971. Austria’s Monetary Policy – Its Methods and Results. In: Schmitz, W. (ed.). Convertibility, Multilateralism and Freedom. Springer. New York. 37–48. Kurier. 1971a. Harter Schilling braucht keine neue Aufwertung. Androsch deutet zwei Dollar-Kurse an. Wednesday. August 18. 1. Kurier. 1971b. Sonntag nach harter Geheimkonferenz. Die Weichen für den Schilling gestellt. Monday. August. 23. 1. Machlup, F. and B. G. Malkiel (eds.). 1964. International Monetary Arrangements. The Prob- lem of Choice. Report on the Deliberations of International Study Group of 32 Economists. Princeton University Press. Princeton. Mooslechner, P., S. W. Schmitz and H. Schuberth. 2007. From Bretton Woods to the Euro: the Evolution of Austrian Monetary Policy from 1969 to 1999. In: From Bretton Woods to the Euro – Austria on the Road to European Integration. Workshops – Proceedings of OeNB Workshops 11. Vienna. 21–44. Nationalbank Act (NBG). 1955. Federal Law Gazette No. 184/1955. (English version: Central Bank Act 1984. OeNB.) Nowotny, E. 2007. Die Hartwährungspolitik und die Liberalisierung des Kapitalverkehrs sowie des Finanzsektors. In: From Bretton Woods to the Euro – Austria on the Road to European Integration. Workshops – Proceedings of OeNB Workshops 11. Vienna. 45–68. OeNB. 1961–1978. Wechselkurspolitik 1961–1978: Internationale und österreichische Wechsel kurspolitik. Carton 1. BHA VI/06.6. OeNB Bank History Archives. OeNB. 1969–1973. Protokolle der Sitzungen des Generalrats (Minutes of the meetings of the General Council) No. 258 and extraordinary meeting on May 10. OeNB Bank History Archives. OeNB. 1971a. Kundmachungen 13/71 und 14/71 (OeNB announcements 13/71 and 14/71). OeNB Bank History Archives. OeNB. 1971b. Protokolle der Direktoriumssitzungen (Minutes of the meetings of the Governing Board) Nos. 778, 779, 781, 782, 792, 794 and 795. OeNB Bank History Archives. OeNB. 1971c. Mitteilungen der Oesterreischischen Nationalbank (Communications of the Oesterreichische Nationalbank). Monthly issues. OeNB Bank History Archives. OeNB. 1971d. Official communication to the IMF. August 23. OeNB. 1971e. Various press statements. OeNB. 1971f. Gedächtnisnotiz über die gemeinsame Sitzung des Direktoriums und des Währungspolitischen Ausschusses am 9. Mai 1971 (Memory minutes of the joint meeting of the Governing Board and the Sub-Committee on Monetary Policy on May 9, 1971). OeNB Bank History Archives.

MONETARY POLICY & THE ECONOMY Q3– Q4/16 209 The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator

Schmitz, W. 1968. Die internationale Währungssituation. Österreichisches Bank-Archiv 16/5. 150–162. Schmitz, W. 1969. Balance of payments and the creation of money – Influence of the net balance of payments on changes in money supply. Lecture given at Harvard University. Cambridge M. A. October 6. BHA I/06.Sc. OeNB Bank History Archives. Schmitz, W. 1970. More Flexibility in Exchange Rates – And in Methods. In: Federal Reserve Bank of St. Louis. March. 11–14. Schmitz, W. 1972a. Bretton Woods: What Changes? Reform of the international monetary system can no longer be delayed. In: Columbia Journal of World Business 7/1. 13–19. Schmitz, W. 1972b. Der Schilling in der Periode flexibler Wechselkurse. In: Weber, W. (ed.). Wirtschaft und Verfassung in Österreich. Festschrift für Franz Korinek. Herder. Wien. 281–290. Schmitz, W. 1973. Gold, SDR’s, Liquidity and Inflation. Hearings before the Joint Economic Committee. Congress of the United States. 93rd Congress. 1st Session Part 2. February 1973. U.S. Government Printing Offices. Washington, D. C. 510–514. Schmitz, W. 1974. Der Schilling in der Periode beweglicher Wechselkurse – bisherige Lösungen und mögliche Alternativen. Wirtschaftspolitische Blätter 2. 150–155. Schmitz, S. W. 2016a. Interview with Heinrich Berg, former expert at the OeNB’s Arbitrage Office and Head of Exchange Market Analysis. March 17. Vienna. Mimeo. Schmitz, S. W. 2016b. Interview with Eduard Hochreiter, former Head of Economic Research at the OeNB. April 20. Vienna. Mimeo. Schmitz, S. W. 2016c. Interview with Heinz Kienzl, former Chief Executive Director and former Vice President of the OeNB. January 28. Vienna. Mimeo. Schmitz, S. W. 2016d. Telephone interviews with Klaus Mündl, former Deputy Chief Executive Director of the OeNB. January 29 and April 28. Mimeo. Schubert, A. and T. Theurl. 1995. The 70-Year History of the Schilling – Leitmotifs and Mile- stones of a Currency. Reports and summaries of the Oesterreichische Nationalbank 1/1995. 41–57. Socher, K. 1973. Österreichische Wechselkurspolitik. Beiträge zur Stabilitätspolitik. Dr. Stigleit- ner Schriftenreihe 13. Österreichisches Forschungsinstitut für Sparkassenwesen 13. 101–117. Socher, K. 1980. The Experience with Floating Exchange Rates in Austria. In: Frisch, H. and G. Schwödiauer (eds.). The Economics of Flexible Exchange Rates. Supplements to Kredit und Kapital 6. 401–410. Socher, K. 1996. Die Entstehung der Hartwährungspolitik. In: Socher, K. (ed.). Wolfgang Schmitz – Wirtschaftspolitische Weichenstellungen 1963–1973. Orac Verlag. Wien. 85–96. Solomon, R. 1977. The International Monetary System 1945–1976. An Insider’s View. Harper and Row. New York. SMP (Subcommittee on Monetary Policy). 1971a. Memo of the Meeting of the Sub committee on Monetary Policy of the General Council of May 9. OeNB Bank History Archives. SMP (Subcommittee on Monetary Policy). 1971b. Minutes of the Meetings of the Sub committee on Monetary Policy of the General Council of August 5. Including annex: Ein “Kapitalverkehrsschilling” in Österrreich? OeNB Bank History Archives. Straumann, T. 2010. Fixed Ideas of Money. Small States and Exchange Rate Regimes in Twentieth-Century Europe. Cambridge University Press. Cambridge. Streissler, E. W. 1998. Neue finanzmarktpolitische Aufgaben des Staates. Conturen II. 7–23. Volkszeitung Kärnten. 1971. Schmitz-Kurs setzt sich durch. Dollarnotierung wird täglich festgesetzt. Nationalbank bleibt am Markt. Wednesday. August 25. 1. Wiener Zeitung. 1971. Österreich wertet nicht auf. Außerordentlicher Ministerrat beriet währungspolitische Lage: Dollar kann unter bisherigen Interventionsbanbreite sinken. Relation zu europäischen Währungen stabil. Tuesday. August 24. 1.

210 OESTERREICHISCHE NATIONALBANK Notes List of studies published in Monetary Policy & the Economy

For further details on the following publications, see www.oenb.at.

Issue Q3/15 Austria: Sluggish economic growth Martin Schneider

Causes of declining investment activity in Austria Gerhard Fenz, Christian Ragacs, Martin Schneider, Klaus Vondra, Walter Waschiczek

Expected retirement age and pension benefits in Austria: evidence from survey data Markus Knell, Esther Segalla, Andrea Weber

Issue Q4/15 Austrian economy to grow at same pace as euro area economy in 2016 and 2017 – economic outlook for Austria from 2015 to 2017 (December 2015) Gerhard Fenz, Martin Schneider

Interest rate perceptions and expectations when interest rates are low – survey evidence on Austrian households Christian Beer, Ernest Gnan, Doris Ritzberger-Grünwald

Financing the Austrian economy – a bird’s eye view based on the financial accounts from 1995 to 2014 and a look at the road ahead Michael Andreasch, Pirmin Fessler, Martin Schürz

Issue Q1/16 Austrian economy fueled by growth stimuli, yet fraught with external risks Gerhard Fenz, Fabio Rumler

Saving, portfolio and loan decisions of households when interest rates are very low – survey evidence for Austrian households Christian Beer, Ernest Gnan, Doris Ritzberger-Grünwald

The Stability and Growth Pact since 2011: More complex – but also stricter and less procyclical? Doris Prammer, Lukas Reiss

Issue Q2/16 Economic recovery in 2016 after four years of weak growth – economic outlook for Austria from 2016 to 2018 (June 2016) Christian Ragacs, Klaus Vondra

Eurosystem Household Finance and Consumption Survey 2014 – first results for Austria (second wave) Pirmin Fessler, Peter Lindner, Martin Schürz

212 OESTERREICHISCHE NATIONALBANK List of studies published in Monetary Policy & the Economy

Issue Q3– Q4/16 A (not so brief) history of inflation in Austria Christian Beer, Ernest Gnan, Maria Teresa Valderrama

The measurement of inflation in Austria: a historical overview Manfred Fluch

Two centuries of currency policy in Austria Heinz Handler

The financial relations between the Nationalbank and the government Doris Prammer, Lukas Reiss, Walpurga Köhler-Töglhofer

Florin, crown, schilling and euro: an overview of 200 years of cash in Austria Clemens Jobst, Helmut Stix

Cashless payments in Austria: the role of the central bank Hans Kernbauer

Principles, circumstances and constraints: the Nationalbank as lender of last resort from 1816 to 1931 Clemens Jobst, Kilian Rieder

The changing role of macroprudential policy in Austria after World War II Sophia Döme, Stefan W. Schmitz, Katharina Steiner, Eva Ubl

The OeNB’s reaction to the end of the Bretton Woods system: tracing the roots of the Indicator Stefan W. Schmitz

MONETARY POLICY & THE ECONOMY Q3– Q4/16 213 Periodical publications

Starting from 2016, the OeNB’s periodical publications are available in electronic format only. They can be downloaded at https://www.oenb.at/en/Publications.html. If you would like to be notified about new issues by e-mail, please register at https://www.oenb.at/ en/Services/Newsletter.html. Geschäftsbericht (Nachhaltigkeitsbericht) German 1 annually Annual Report (Sustainability Report) English 1 annually This report informs readers about the Eurosystem’s monetary policy and underlying economic conditions as well as about the OeNB’s role in maintaining price stability and financial stability. It also provides a brief account of the key activities of the OeNB’s core business areas. The OeNB’s financial statements are an integral part of the report. http://www.oenb.at/en/Publications/Oesterreichische-Nationalbank/Annual-Report.html

Inflation aktuell German 1 quarterly This publication presents the OeNB’s analysis of recent inflation developments in Austria and its inflation outlook for Austria for the current and next year. In addition, it provides in-depth analyses of topical issues.

Konjunktur aktuell German 1 seven times a year This publication provides a concise assessment of current cyclical and financial developments in the global economy, the euro area, Central, Eastern and Southeastern European countries, and in Austria. The quarterly releases (March, June, September and December) also include short analyses of economic and monetary policy issues. http://www.oenb.at/Publikationen/Volkswirtschaft/Konjunktur-aktuell.html

Monetary Policy & the Economy English 1 quarterly This publication assesses cyclical developments in Austria and presents the OeNB’s regular macro economic forecasts for the Austrian economy. It contains economic analyses and studies with a particular relevance for central banking and summarizes findings from macroeconomic workshops and conferences organized by the OeNB. http://www.oenb.at/en/Publications/Economics/Monetary-Policy-and-the-Economy.html

Fakten zu Österreich und seinen Banken German 1 twice a year Facts on Austria and Its Banks English 1 twice a year This publication provides a snapshot of the Austrian economy based on a range of structural data and indicators for the real economy and the banking sector. Comparative international measures enable readers to put the information into perspective. http://www.oenb.at/en/Publications/Financial-Market/Facts-on-Austria-and-Its-Banks.html

Financial Stability Report English 1 twice a year The reports section of this publication analyzes and assesses the stability of the Austrian financial system as well as developments that are relevant for financial stability in Austria and at the international level. The special topics section provides analyses and studies on specific financial stability-related issues. h t t p://w ww.oenb.at/en/Publications/Financial-Market/Financial-Stability-Report.html

Focus on European Economic Integration English 1 quarterly This publication presents economic analyses and outlooks as well as analytical studies on macroeco nomic and macrofinancial issues with a regional focus on Central, Eastern and Southeastern Europe. http://www.oenb.at/en/Publications/Economics/Focus-on-European-Economic-Integration.html

214 OESTERREICHISCHE NATIONALBANK Periodical publications

Statistiken – Daten & Analysen German 1 quarterly This publication contains analyses of the balance sheets of Austrian financial institutions, flow-of- funds statistics as well as external statistics (English summaries are provided). A set of 14 tables (also available on the OeNB’s website) provides information about key financial and macroeconomic indicators. http://www.oenb.at/Publikationen/Statistik/Statistiken---Daten-und-Analysen.html

Statistiken – Daten & Analysen: Sonderhefte German 1 irregularly Statistiken – Daten & Analysen: Special Issues English 1 irregularly In addition to the regular issues of the quarterly statistical series “Statistiken – Daten & Analysen,” the OeNB publishes a number of special issues on selected statistics topics (e.g. sector accounts, foreign direct investment and trade in services). http://www.oenb.at/en/Publications/Statistics/Special-Issues.html

Research Update English 1 quarterly This newsletter informs international readers about selected research findings and activities of the OeNB’s Economic Analysis and Research Department. It offers information about current publications, research priorities, events, conferences, lectures and workshops. Subscribe to the newsletter at: http://www.oenb.at/en/Publications/Economics/research-update.html

CESEE Research Update English 1 quarterly This online newsletter informs readers about research priorities, publications as well as past and upcoming events with a regional focus on Central, Eastern and Southeastern Europe. Subscribe to the newsletter at: http://www.oenb.at/en/Publications/Economics/CESEE-Research-Update.html

OeNB Workshops Proceedings German, English 1 irregularly This series, launched in 2004, documents contributions to OeNB workshops with Austrian and international experts (policymakers, industry experts, academics and media representatives) on monetary and economic policymaking-related topics. http://www.oenb.at/en/Publications/Economics/Proceedings-of-OeNB-Workshops.html

Working Papers English 1 irregularly This series provides a platform for discussing and disseminating economic papers and research findings. All contributions are subject to international peer review. http://www.oenb.at/en/Publications/Economics/Working-Papers.html

Proceedings of the Economics Conference English 1 annually The OeNB’s annual Economics Conference provides an international platform where central bankers, economic policymakers, financial market agents as well as scholars and academics exchange views and information on monetary, economic and financial policy issues. The proceedings serve to document the conference contributions. http://www.oenb.at/en/Publications/Economics/Economics-Conference.html Proceedings of the Conference on European Economic Integration English 1 annually The OeNB’s annual Conference on European Economic Integration (CEEI) deals with current issues with a particular relevance for central banking in the context of convergence in Central, Eastern and Southeastern Europe as well as the EU enlargement and integration process. For an overview see: http://www.oenb.at/en/Publications/Economics/Conference-on-European-Economic-Integration-CEEI.html The proceedings have been published with Edward Elgar Publishers, Cheltenham/UK, Northampton/ MA, since the CEEI 2001. www.e-elgar.com

Publications on banking supervisory issues German, English 1 irregularly http://www.oenb.at/en/Publications/Financial-Market/Publications-of-Banking-Supervision.html

MONETARY POLICY & THE ECONOMY Q3– Q4/16 215 Addresses

Postal address Phone/fax/e-mail

Head office Otto-Wagner-Platz 3 PO Box 61 Phone: (+43-1) 404 20-6666 1090 Vienna, Austria 1011 Vienna, Austria Fax: (+43-1) 404 20-042399 Internet: www.oenb.at E-mail: [email protected]

Branch offices Northern Austria Branch Office Coulinstraße 28 PO Box 346 Phone: (+43-732) 65 26 11-0 4020 Linz, Austria 4021 Linz, Austria Fax: (+43-732) 65 26 11-046399 E-mail: [email protected]

Southern Austria Branch Office Brockmanngasse 84 PO Box 8 Phone: (+43-316) 81 81 81-0 8010 Graz, Austria 8018 Graz, Austria Fax: (+43-316) 81 81 81-046799 E-mail: [email protected]

Western Austria Branch Office Adamgasse 2 Adamgasse 2 Phone: (+43-512) 908 100-0 6020 Innsbruck, Austria 6020 Innsbruck, Austria Fax: (+43-512) 908 100-046599 E-mail: [email protected]

Representative offices New York Representative Office Phone: (+1-212) 888-2334 Oesterreichische Nationalbank Fax: (+1-212) 888-2515 450 Park Avenue, Suite 1202 10022 New York, U.S.A.

Brussels Representative Office Phone: (+32-2) 285 48-41, 42, 43 Oesterreichische Nationalbank Fax: (+32-2) 285 48-48 Permanent Representation of Austria to the EU Avenue de Cortenbergh 30 1040 Brussels, Belgium

216 OESTERREICHISCHE NATIONALBANK