American Economic Association

Stabilization Policies in the World Economy: Scope and Skepticism Author(s): Reviewed work(s): Source: The American Economic Review, Vol. 72, No. 2, Papers and Proceedings of the Ninety- Fourth Annual Meeting of the American Economic Association (May, 1982), pp. 56-61 Published by: American Economic Association Stable URL: http://www.jstor.org/stable/1802303 . Accessed: 20/04/2012 13:46

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

American Economic Association is collaborating with JSTOR to digitize, preserve and extend access to The American Economic Review.

http://www.jstor.org Stabilization Policies in the World Economy: Scope and Skepticism

By JEFFREY SACHS*

Throughout the industrialized world, mac- ills, the sources of relative success across roeconomic performance since the mid- 1970's economies, or most important for the pur- has been very poor, and the prospects in the pose here, the right policy mix for sustained near term remain bleak. It is sobering to recovery. The interpretations offered here, reflect that all twenty-four OECD economies which must be regarded as tentative, lay suffered a slowdown in aggregate economic great stress on the various adverse supply growth after 1973 (comparing average growth shocks that affected all of the OECD econo- rates for 1965-73 with 1973-79); all but one mies during the past decade. These interpre- (Switzerland) experienced an intensification tations are based largely on a joint research of consumer price inflation. Overall, the an- project with Michael Bruno on the macro- nual GNP growth of the OECD slowed from economics of supply disturbances in open 4.9 percent during 1965-73 to 2.7 percent economies. during 1973-79, and it has slowed further since then. The slow growth has translated I. The CentralRole of SupplyDisturbances into rising unemployment, which stood at about 7 percent of the OECD workforce in The major competing schools of macro- 1981 as compared with a mere 2 to 3 percent economic thought have focused most of their in 1970. In the European Economic Com- blame for the current debacle on macroeco- munity, the 1981 unemployment rate ap- nomic policy. For the Keynesians, recent pears to have been a staggering 8 percent. policy has been too austere, overly directed Bright spots in this picture are few indeed, against fighting inflation. For the monetar- but their lessons may be instructive. After ists, the case has been almost the opposite: the steep recession in the OECD during that politicians have continued to drive up 1974-75, the United States alone of the major money growth to fight short-term unemploy- economies staged a rapid recovery; the un- ment, to the sacrifice of longer-term price employment rate fell significantly below its stability. And for the new classical macro- 1975 peak, while it continued to rise in economists, the policies have simply been too Europe. Unfortunately, U.S. performance on erratic, with policy " surprises" explaining most measures has deteriorated sharply (and the fluctuations in output growth. relatively) since 1979. Japan provides a case Unfortunately, these important proposi- of extremely successful performance since tions have been subjected to almost no sys- the mid-1970's, after a very sharp jolt during tematic, cross-country scrutiny. No strong 1973-75. Among the smaller economies, the comparative evidence has been set forth to neighboring economies of Sweden and Fin- show that high unemployment and slow land offer a vivid contrast of worsening and growth have been closely tied to more re- improving economic developments, respec- strictive policies, or to more uncertain and tively, since the mid-1970's. volatile policies, or that price stability has There is no consensus among macro- resulted from slow and stable money growth. economists regarding the diagnosis of these There is little doubt that tight policies can explain high unemployment at certain times * and National Bureau of Eco- and places (the Thatcher experiment in the nomics. This work is part of a project with Michael U.K. is a case in point; see Willem Buiter and Bruno of Hebrew University on the of supply shocks. Errors in this analysis are, of course, my Marcus Miller), but it is doubtful that they own. Support from the National Science Foundation is provide a general explanation for the recent gratefully acknowledged. experience. The almost universal slowdown

56 VOL. 72 NO.2 MACROECONOMIC POLICY IN THE OPEN ECONOMY 57 in growth and rise in unemployment in the steel, shipbuilding, electronic components). OECD has characterized both activist and This import-competing growth of the NICs passive, as well as expansionary and contrac- has worsened the OECD terms of trade, and tionary policy regimes. To cite just two cases, perhaps more importantly, has shifted the neither the austere Barre policies in France locus of new world investment in key in- nor the expansionary Keynesian policies in dustries decisively away from the developed Sweden restored high employment or rapid economies. growth to their economies. On a theoretical level, the economics of I would suggest three lessons on macro- supply shocks are fairly well understood (see economic performance and policy from the my papers with Bruno 1981a, b, and my tangled comparative record. First, it is not 1982 article for general equilibrium analyses the policy choices but rather the policy op- in the case of output-market clearing, and tions that worsened in the 1970's, with supply and for shocks driving the stagflationary process. the non-market-clearing case). Consider, for Second, the appropriate policy response to example, a rise in real input prices. In a high unemployment or slow growth depends competitive, full-employment economy a on the source of the unemployment, with permanent input price shock reduces output "'supply-generated"unemployment less trac- on impact, and most likely sets in motion a table then the garden-variety Keynesian un- path of capital decumulation, along which employment. And third, since national eco- output and productivity grow more slowly nomic structures differ, particularly in labor than trend. For a given money supply, the market and financial institutions, the same nominal price and wage levels may either rise policy is likely to have very different effects or fall after the shock, though the shock across economies. probably requires the fall of other nominal Various "supply-side" shocks were of wages and prices. And very importantly, the dominating importance in the 1970's. All real wage consistent with full employment industrial countries faced a massive rise in (hereafter, the "full-employment real wage") the real price of raw materials inputs (both must fall in impact, and then must grow fuel and nonfuel) after 1970, following two more slowly than trend as the process of decades of falling real input prices. The oil capital accumulation proceeds. price increases (in 1973-74, 1979-80) were From the point of view of macroeconomic the most stunning, but by no means the only equilibrium, then, there are two problems. major hikes in real commodity prices. Over- After the supply shock, the nominal price all, the index of nonfuel primary input prices vector may be inappropriate given the exist- rose sharply relative to prices of final manu- ing money stock and exchange rate. If nomi- factured goods in the 1970's (see Irving nal prices and/or wages are sticky, a stan- Kravis and Robert Lipsey). Accompanying dard demand management problem arises these shocks was a persistent slowdown in (see Edmund Phelps in this context). The total factor productivity growth in almost all second and more novel policy issue involves of the OECD. The synchronization of the the need to reduce real wages to their new slowdown with the raw material price in- full-employment path. Most recessions up creases suggests a causal linkage running until 1973 signalled little about the need for from prices to output (as suggested by Martin long-term real wage adjustment, while the N. Baily and Bruno, for example) but this resolution of post-1973 recessions has de- linkage remains an open question. Many, pended on the deceleration of real wages such as David Grubb, Richard Jackman, from an earlier trend. And for reasons that Richard Layard, and William Nordhaus, take we shall see, such a deceleration is only likely the productivity slowdown to be an indepen- after a transitory phase of high unemploy- dent event. A third supply shock, from the ment, and is also likely to be hard to bring point of view of the OECD, has been the about with standard macroeconomic policy rapid expansion of the newly industrializing tools. countries (NIC) into traditional export sec- For a number of years after 1973 (at least tors of the OECD economies (for example, two years in most countries: four to five in 58 A FA PA PERS A ND PROCEEDINGS MA Y 1982

others) OECD real wage growth remained A related argument holds that unions, or strong relative to productivity growth, and both employers and unions, failed to under- profitability was sharply squeezed (see my stand the link of higher oil prices and wage 1979 article, and Bruno's paper for details). moderation, and the ordeal of unemploy- The rate of return on manufacturing capital ment was necessary to "clarify" that link. fell steeply between 1973 and 1978 in most This simple argument probably holds enor- economies. Only in the United States, where mous weight. The supply shocks were a novel unionization rates are extraordinarily low, phenomenon. There was no way prior to the was the profit squeeze largely avoided. The late 1970's to evaluate the partial elasticity of evidence is not strong on whether the real labor demand with respect to real energy wage behavior reflects union wage setting prices, or to verify that a persistent slow- per se, or more general outcomes in the labor down in productivity growth had occurred. market. It is significant though, that in both An asseveration by employers to employees the United Kingdom and United States there of the need for real wage declines is, in was a sharp rise in the union-nonunion wage general, of little avail, for employers have differential over the course of the 1970's. reason to dissimulate and employees have The profit squeeze was closely linked to cause to ignore their employer's importun- output, investment, and growth behavior ings. The inability of employers to convey after 1973, with the relatively favorable U.S. credibly to workers the need for real wage profit position inducing a more rapid re- moderation has been elegantly captured by covery (see my 1979 article). The links of Sanford Grossman and Oliver Hart. Adverse wages to unemployment in this period are shocks in their model bump against partial best documented for the U.K. (see James real wage rigidity, and therefore cause unem- Symons for a detailed presentation; R. Mor- ployment. Their results directly transfer to ley; my 1981b paper with Bruno), and our case. If this hypothesis is maintained, we econometric work supports this link for Japan should expect to see a gradual process of (see my paper with David Lipton). Indeed, wage moderation after a supply shock, as Japan provides a revealing comparison of workers gain evidence (through the per- adjustment to the first and second oil shocks. sistence of unemployment) that the adverse In the first, real wage growth remained demand shift against labor has in fact oc- high, and profits and output were sharply curred. Moreover, we would expect learning squeezed; in the second, there was a dramatic between the first (1973-74) and second oil drop in real wage growth for 1979-80, which shocks (1979) regarding union wage setting. made room for Japan's terms-of-trade loss. According to Shinkai, Japan is a vivid illus- Output growth hardly dipped in the second tration of such learning, for formerly mili- episode (see my 1981 paper, and Yoichi tant union federations explicitly called for Shinkai for supporting evidence). wage moderation in 1980, in light of supply- Real wages may fail to adjust for many side developments. reasons, and each of these reasons has differ- The previous explanations all apply in a ent implications for policy. We can divide basically competitive labor market setting. the possible explanations into categories More troubling cases emerge once we recog- which emphasize: uncertainty, timing, or nize the extent of monopoly union power misperception; and union bargaining power. in OECD labor markets, particularly Most directly, some wages may be prede- throughout Europe. In the United States, we termined by contract at the time of an unan- often forget that much of European wage ticipated supply disturbance, so that real setting occurs in a highly centralized, highly wages are unexpectedly driven above full- unionized context. And when powerful un- employment levels. If renegotiation is cost- ions face off against employers, supply shocks less, the profit squeeze would soon disap- may well redound on unemployment rather pear, but otherwise the squeeze must persist than wage reductions. To our benefit, Ian until the next bargaining round. McDonald and Solow have recently offered VOL. 72 NO.2 MACROECONOMIC POLICY IN THE OPEN ECONOMY a smorgasbord of models that make that very occur with high or low real-wage bargains. If point. There is simply no presumption that W>Wf-WB, unions and firms would settle an optimizing union will substantially cut on a lower real wage, but because of partial real wages, rather than employment, follow- nominal wage rigidity, they do not achieve it. ing a supply shock; indeed, it may even raise This is a case where a money supply increase them! or exchange-rate devaluation can readily re- duce unemployment (at the cost of a higher II. Implicationsfor DemandManagement price level). Unemployment is basically a Policiesin OpenEconomies monetary problem. On the other hand, if w = wB > wf, the From the very aggregative standpoint then, wage bargain is intentionally set above wf, supply shocks may raise the "typical" prob- (as discussed earlier, unions may misperceive lem that the nominal price and wage are out wf or may choose unemployment in return of line with money supply and the exchange for higher wages). In this case, a monetary rate, and the novel case that real wages ex- expansion can temporarily reduce w, and ceed their full employment level. If output increase employment, but only in the short markets do not clear, we may adopt run. Long-run, full-employment equilibrium Malinvaud's typology: the first problem requires that wB be reduced to wf (or that wf would push the economy into the regime of be raised). For concreteness, suppose that Keynesian unemployment, and the latter to- AwB = a(U,), where U is the unemployment wards classical unemployment. Demand rate. Then, 1/a(wB - wf ) measures the stimulus is effective in the first case (subject cumulative unemployment that must be ex- to Mundell-Fleming limitations) but not in perienced before long-run equilibrium is re- the latter, unless the demand stimulus itself stored. The unemployment may be post- (say an exchange-rate depreciation) can re- poned through rising inflation, but it cannot duce real wages by accelerating inflation. be avoided in the long run through expan- In recent models (particularly my 1982 sionary monetary policy. Simulation ex- article, and my 1981a article with Bruno) I ercises show that expansionary policy very have investigated these policy issues under a often results in higher inflation and deeper variety of labor market assumptions (in these unemployment along the adjustment path models, the output market is assumed to than does a passive policy. clear continuously, so that the economy is It should be pointed out that a fiscal ex- always at the boundary of Keynesian and pansion may raise wf and thus moderate classical regimes). Consider one useful speci- unemployment by favorably shifting the fication, which distinguishes between the economy's terms of trade. Moreover, direct "bargained" real wage w B, the actual real supply-side measures may also raise wf. Space wage w, and the full-employment real wage prevents elaboration of these two possibil- wf (see Grubb et al. for a similar formula- ities here. tion). We assume that unions and firms I close with a tale of two countries, Sweden bargain for a real wage (w B ), but set a and Finland, that vividly confirms the diffi- nominal wage that is only partially indexed culty of preventing, rather than merely post- (or perhaps fully indexed with a lag). Actual poning, supply-shock unemployment.' Swe- wages (w) can deviate in the short run from den utilized very expansionary policies dur- wB, because of unanticipated or accelerating ing 1974-76 to "bridge" the world recession inflation. The bargained wage itself is as- sumed to respond only to in unemployment, 'The year-to-year GNP developments in the two order to capture the partial real wage rigidity countries in the mid- 1970's were discussed above. Now, we can envision various relations 1973 1974 1975 1976 1977 1978 1979 among wB, w, and wf Generally, unemploy- Finland 6.5 3.1 0.6 0.3 0.3 2.3 7.2 ment will result when w> wf, but this can Sweden 3.8 4.2 2.5 1.6 -2.4 1.3 4.1 60 A EA PA PERS A ND PROCEEDINGS MA Y 1982

(see OECD Economic Surveys for Sweden REFERENCES and Finland for recent policy history), and indeed open unemployment did not develop Baily, Martin N., " Productivity and the in the early years of the policy (though hid- Services of Capital and Labor," Brookings den unemployment, such as workers in Papers on Economic Activity, Washington government job-training programs, did). But 1981, No. 1, 1-50. neither did real wages moderate. A wage Bruno,Michael, "Raw Materials, Profits, and boom and severe profit squeeze ensued, which the Productivity Slowdown," working ushered in a number of years of very poor paper no. 660, National Bureau of Eco- growth. Moreover, the expansionary policies nomic Research, December 1981. left a legacy of a greatly expanded govern- _ and Sachs, Jeffrey, (1981 a) "Supply ment sector. In Finland, the decision to Versus Demand Approaches to the Prob- abandon reflation was taken much earlier lem of Stagflation," in H. Giersch, ed., (1975), with a view towards restoring profit- Macroeconomic Policies for Growth and ability and competitiveness in Finland's ex- Stability, Kiel: Institute fur Weltwirtscheft port sector. The output drop in Finland in an der Universitat Kiel, 1981. 1975-76 was far more severe than in Sweden, and , (1981b) "Theory of an and Finland's real wage gap (as measured by Input Price Shock and the Dynamic Ad- the OECD) was eliminated in 1976. The justment Process," presented at the Con- growth since 1977 has been far higher. ference in Unemployment, Newnham Col- lege, Cambridge, 1981. III. Conclusions and Extensions Buiter, Willem and Miller, Marcus, "The Thatcher Experiment - What Went The recessions in the 1970's were inher- Wrong?,"Brookings Papers on Economic ently more painful than previous episodes Activity, forthcoming. since they signalled the need for real wage Grossman,Sanford, and Hart, Oliver, "Implicit moderation and a period of slow economic Contracts, Moral Hazard, and Unemploy- growth. In some countries, the need for real ment," American Economic Review Pro- wage moderation was accepted by workers ceedings, May 1981, 71, 301-07. without the ordeal of unemployment (for Grubb,D., Jackman,R., and Layard,R., "Causes example, Japan, 1979-80), while in others, of the Current Stagflation," Centre for the adjustment process seemed to require a Labour Economics discussion paper no. recession. In such countries, expansionary 96, London School of Economics, 1981. demand policies serve mainly to postpone Kravis, I. B., and Lipsey, R. E., "Prices and rather than prevent an economic downturn. Terms of Trade for Developed Country Of course, a variety of additional issues Exports of Manufactured Goods," work- should be raised in a complete treatment of ing paper no. 774, National Bureau of the recent supply shocks. Higher energy Economic Research, September 1981. prices and competition from the NICs call Lipton,David and Sachs, Jeffrey, "The Supply for sectoral reallocation of resources, as well Approach to Oil Shocks and the Slow- as overall wage and price adjustments. One down in Japanese Economic Growth," suspects that in a booming economy, sectoral Oriental Economist, 1981. shifts of the requisite magnitude could be McDonald, Ian and Solow, Robert, "Wage handled in stride. In the present environ- Bargaining and Employment," American ment, though, policymakers have supported Economic Review, December 1981, 71, moribund industries and protective labor 896-908. legislation has slowed down the flow of Malinvaud,Edmond, Profitability and Unem- workers to more productive enterprises. ployment, Cambridge: Cambridge Univer- These inefficiencies have magnified Europe's sity Press, 1980. adjustment problem significantly. Morley, R., "Profit, Relative Prices and Un- VOL. 72 NO. 2 MA CROECONOMIC POLICY IN THE OPEN ECONOMY 61

employment," Economic Journal, Septem- Japanese Ministry of Finance, Conference ber 1979, 89, 582-600. on Macroeconomics, Tokyo, August, 1981. Nordhaus,William, "Economic Policy in the . "Energy and Growth Under Flexi- Face of Declining Productivity Growth," ble Exchange Rates," in J. Bhandhari and European Economic Review, 1982 forth- B. Putnam, eds., The International Trans- coming. mission of Economic Disturbances Under Phelps, Edmund,"Commodity Supply Shock Flexible Exchange Rates, Cambridge: MIT and Full Employment Monetary Policy," Press, 1982. Journal of Money, Credit, and Banking, Shinkai, Yoichi, "Oil Crises and the Stag- May 1978, 10, 206-21. flation (or Its Absence) in Japan," discus- Sachs, Jeffrey, "Wages, Profits, and Macro- sion paper no. 110, Institute of Social and economic Adjustment: A Comparative Economics Research, Osaka University, Study," Brookings Papers on Economic Ac- November 1980. tivity, Washington 1979, No. 2, 269-319. Solow, Robert, "What to do When OPEC ,"Wages, Flexible Exchange Rates, Comes," in , ed., Rational and Macroeconomic Policy," Quarterly Expectations and Economic Policy, Chi- Journal of Economics, June 1980, 94, 737- cago: University of Chicago Press, 1980. 47. Symons,James, "The Demand for Labour in , "The Macroeconomic Response to British Manufacturing," Centre for Labour Supply Shocks: A Comparison of Japan Economics discussion paper no. 91, and the United States," presented at the London School of Economics, 1981.