ECON2915 Economic Growth Lecture 1 : Introduction to economic growth
Andreas Moxnes
University of Oslo
Fall 2016
1 / 41 This Term
13 lectures - every Wednesday (with exceptions). 10 seminars (starts two weeks after 1st lecture). Term paper, must pass in order to take exam. 3 hour exam Dec 16. Grade based on exam but participation in class highly encouraged. Contact student: (i) Contact between me and you, (ii) participate in evaluating the course.
2 / 41 The Syllabus
Economic growth:
I Weil, David: Economic Growth (3rd edition). 2013. Chapters 1-4 and 6-13. International linkages:
I McLaren, John: International Trade. 2012. Chapters 1-2, 5-6, 9. Various papers (recommended reading).
3 / 41 Tentative Schedule
Lecture 1. Introduction to economic growth (W 1-2). Lecture 2. The Solow model (W 3). Lecture 3. Malthus. Population growth (W 4). Lecture 4. Human capital and productivity measurement (W 6-7). Lecture 5. Technology (W 8-9). Lecture 6. Efficiency (W 10). Lecture 7. Institutions (W 12). Lecture 8. Growth in the open economy. (W 11, MC 1). Lecture 9. The Ricardian model (MC 2). Lecture 10. Specific factors (MC 5). Lecture 11. The Hecksher Ohlin model (MC 6). Lecture 12. Trade policy and growth (MC 9). Lecture 13. Summing up.
4 / 41 Introduction to Economic Growth
Why study economic growth?
I Large differences in living standards across countries. I Why are some countries rich and others poor? Large variation over time within countries.
I Growth miracles like China. I How can we improve growth in poor countries?
5 / 41 How to measure economic well-being
Gross Domestic Product (GDP) : The value of all goods and services produced in a country in a given year.
I GDP per capita : Income/output per person. Can compare GDP across countries and over time. Can you think of other measures of (economic) well-being?
6 / 41 Top 11 countries, 2009
Some overlap between columns 2 and 3 but not 1.
7 / 41 Enormous differences across countries
The top 20% of world population has income > average. Top 5% of world population has income > $40,000
8 / 41 THE FACTS OF ECONOMIC GROWTH 37
The Great DivergenceFigure 21:TheGreatDivergence GDP PER PERSON (MULTIPLE OF 300 DOLLARS) 100 U.S.
80 U.K. Japan
60
40 Argentina China 20
Ghana
1200 1300 1400 1500 1600 1700 1800 1900 2000 YEAR
Note: The graph shows GDP per person for various countries, normalized by the value in the United Kingdom in the initial year. Source: The Maddison Project, Bolt and van Zanden (2014). Small income differences prior to year 1600. Ratio of richest to poorest GDP/capita around 5. Today the ratio is > 100 (US vs Ghana) 9 / 41 THE FACTS OF ECONOMIC GROWTH 39
Convergence?Figure 23:TheSpreadofEconomicGrowthsince1980 GDP PER PERSON (US=100)
United States 80 Japan Western Europe 40 Russia Brazil 20
China 10 India
5
Sub-Saharan Africa
1980 1985 1990 1995 2000 2005 2010 YEAR China and India, YES. Sub-Saharan Africa, NO. Source: The Penn World Tables 8.0. In general, catch-up among OECD countries (middle income countries). But not for poor countries.
I E.g. yes for Botswana and South Korea but not for Madagascar and Niger 4.2. The Spread of Growth in Recent Decades 10 / 41
Figure 23 focuses in on the last 30 years using the Penn World Table 8.0 data, again showing GDP per person relative to the U.S. Several facts then stand out. First, incomes in the countries of Western Europe have been roughly stable, around 75 percent of the U.S. level. It is perhaps surprising that countries like France, Germany, and the U.K. are this far behind the United States. Prescott (2004) observes that a large part of the difference is in hours worked: GDP per hour is much more similar in these countries, and it is the fact that work hours per adult are substantially lower in Western Europe that explains their lower GDP per person. Jones and Klenow (2015) note that in addi- tion to the higher leisure, Western Europeans tend to have higher life expectancy and lower consumption inequality. Taking all of these factors into account in constructing a consumption-equivalent welfare measure, the Western European countries look much closer to U.S. levels than the simple GDP per person numbers imply. Figure 23 also illustrates the “lost decades” that Japan has experienced. After rapid growth in the 1980s (and before), Japan peaked at an income relative to the U.S. of 85 Control for differences in purchasing power.
I Use PPP exchange rates (purchasing power parity).
GDP comparisons
Over time: Control for inflation (“deflated GDP”) Across countries: Converting GDP to common currency (e.g. USD) using nominal exchange rate problematic.
11 / 41 GDP comparisons
Over time: Control for inflation (“deflated GDP”) Across countries: Converting GDP to common currency (e.g. USD) using nominal exchange rate problematic.
Control for differences in purchasing power.
I Use PPP exchange rates (purchasing power parity).
11 / 41 Purchasing Power Parity
Why not compare GDP measured in the same currency (e.g., USD)?
I The price level very different in rich and poor countries.
F Non-traded goods are typically cheaper in poor countries. F $1 has higher purchasing power in poor countries.
I Currencies are volatile.
12 / 41 PPP
PPP exchange rates calculated by measuring the cost of a basket of goods and services.
I E.g. basket = (1 TV, 10 haircuts). Price of basket is 1×10+10×2=30 in R, 1×10+10×1=20 in P. PPP exchange rate: R-dollar=2/3 P dollar −→ P dollar worth more. Relative GDP using exchange rates: 120/20=6 (1-1 ex. rate). Relative GDP using PPP exchange rates: 120R 120 2 2 = = 6 = 4. 20P 20 3 3
13 / 41 PPP
The effect of using PPP comparisons of GDP:
14 / 41 Example: The Big Mac Index
15 / 41
Pros and cons of this PPP exchange rate? Also used to gauge whether a currency is over/undervalued. Example: The Big Mac Index
16 / 41 Example: The Big Mac Index
The Penn effect / Balassa-Samuelson effect.
17 / 41 From levels to growth
Why is growth important? Annual average U.S. growth (per capita GDP) during 1870-2009 was 1.8%. Small’ish number but nevertheless dramatic changes over the course of a century. The power of compound growth: 1.018139 ≈ 12. But high sustained growth a recent phenomenon.
I Living standards doubled from year 1 to 1820 −→ annual growth of 0.04%.
18 / 41 GDP per capita in the U.S.
.
19 / 41 Log GDP per capita in the U.S.
Great Depression barely noticeable in the long run.
20 / 41 Log GDP per capita in Norway
100000
10000 1865 1868 1871 1877 1889 1895 1898 1901 1907 1919 1925 1928 1931 1937 1949 1955 1958 1961 1967 1979 1985 1988 1991 1997 2009 2000 2003 2006 1874 1880 1883 1886 1892 1904 1910 1913 1916 1922 1934 1940 1943 1946 1952 1964 1970 1973 1976 1982 1994
Fixed 2005 NOK prices. From 20,000 NOK in 1870s to 420,000 in 2011. 21 / 41 GDP per capita US/UK/Japan
US 1.8% vs UK 1.5% growth. UK was 31% richer in 1870, 19% poorer today.
22 / 41 The Distribution of Growth Rates, 1975–2009 “Growth miracles” China and Equatorial Guinea vs “growth disasters” Zimbabwe and Liberia
23 / 41 Growth before 1970 Poor data before 1970, but can look at regions instead of countries.
1 Growth has increased over time. 2 Differences between rich and poor countries have increased over time. 24 / 41 Economic growth as seen from outer space Taken at night on Jan 30 2014 from the International Space Station (ISS).
25 / 41 Economic growth as seen from outer space 1002 THE AMERICAN ECONOMIC REVIEW APRIL 2012 1992 & 2008 snapshots. South: Real GDP increase 119%. North: 0?.
1992 2008
Digital Number High : 63
Low : 0
0 25 50 100 km
Universal Transverse Mercator projection Figure 2: Long term growth: Korean peninsula Universal Transverse Mercator projection Lights reflect long termF growthigure 2. Long (Henderson,-Term Growth: Korean Storeygard, Peninsula Weil, 2012). 26 / 41 Source: See Figure 1.
72 percent in the same time period. We don’t expect the percentage growth in income and lights to be the same, both because the elasticity may not be one and because the lights measures were done by different satellites in 1992 and 2008, the sensor settings of which will not exactly match. Offshore lights near South Korea in 1992 are from fshing boats shining bright lights to attract photophilic creatures like squid. Figure 2 also shows the dismal comparative situation in North Korea, with little or no growth in the same time period. The average digital number fell by 7.4 percent.
Indonesia.—To illustrate the high-frequency response of lights to an economic downturn, we use data from Indonesia in 1997, before the Asian fnancial crisis, and in 1998, when Indonesia was at a GDP low. Overall for Indonesia the digital number declined by 6 percent from 1997 to 1998 and real GDP declined by 13 percent. To improve visualization we focus on just the main island of Java, pictured in Figure 3. In Figure 3, lights in 1997 are in the top panel and lights in 1998 are in the second. The third panel shows pixels for which the digital number changed by more than three. There are large patches of declines in lights in west Java, around Jakarta and its suburban areas, and in east Java, around the growth pole of Surabaya and its hinterlands, going southwest from Surabaya. Although declines in lights out- put dominate, in some rural areas there is an increase in lights. We know that there was some return to rural areas by urban migrants in the crisis and that there is also drilling and refning of petroleum in some of these areas. In the bottom panel, we Economic growth as seen from outer space Large crisis event: Rwanda genocide. Sharp fall in GDP from 1993-1994. 1004 THE AMERICAN ECONOMIC REVIEW APRIL 2012
1993 1994
Digital Number High : 63
0 10 20 km Low : 0
Universal Transverse Mercator projection
1996 Advantages:27.5 (1) Not necessary to use country as unit of analysis, (2) Can measure growth for countries with poor national income accounts. Cons? 27 27 / 41
26.5
Actual ln(GDP) ln(GDP) predicted by lights 26 1992 1994 1996 1998 2000 2002 2004 2006 2008 Year Figure 4: Genocide event: Rwanda
Figure 4. Genocide Event: Rwanda
Note: Predicted income is based on the results in Table 2 column 1. Source: See Figure 1.
Rwandan Genocide.—To illustrate how a large crisis event is refected in lights, Figure 4 examines the Rwandan genocide. The lights clearly show a sharp tem- porary dimming from 1993 to 1994, with a return to 1993 levels by 1996. This is visible for the capital Kigali as well as more minor urban centers. The graph in the fgure shows offcially measured GDP along with the level of GDP implied by the lights data from the specifcation in Section III. We note in both Figures 3 and 4 lights underpredict the extent of measured income declines. For Indonesia, where national income data are relatively good, this could be underprediction of the true income decline. For Rwanda, national income data are less reliable and economic activity may have been poorly recorded in the period of genocide. These examples raise the possibility that lights respond asymmetrically to income changes, dimming less in downturns than they rise in periods of growth. In Section III we look explicitly at a form of generalized ratchet effects but reject them. It still may be the case, however, that lights respond sluggishly to short-term economic fuctuations, perhaps because lights are produced by durable goods. We believe lights data are best suited to predicting long-term growth and that is the focus of applications later in the paper. Analytical Framework
Why are some countries rich and others poor? (Weil, ch2) We distinguish between Factor accumulation (labor and capital) Productivity (lectures 2-4).
I Technology and efficiency (lectures 4-6) I Government and institutions (lecture 7).
F Openness and international trade.
28 / 41 Macro Production Function
Relationship between inputs/worker (e.g., capital) and output/worker. 29 / 41 Productivity or factor accumulation
30 / 41 What can we learn from data? Example: Is population growth good or bad for economic growth?
Correlation coefficient = -0.42 : High population growth associated with low income.
Outliers: Bahrain & Saudi. 31 / 41 No.
I Population growth −→ Income OR I Income −→ Population growth OR I A third variable causes both low income & high population growth (omitted variables).
F E.g. health outcomes.
Causality and correlation
Does negative correlation mean that low population growth causes high income?
32 / 41 Causality and correlation
Does negative correlation mean that low population growth causes high income?
No.
I Population growth −→ Income OR I Income −→ Population growth OR I A third variable causes both low income & high population growth (omitted variables).
F E.g. health outcomes.
32 / 41 Identifying causal relationships
Regression analysis (e.g. controlling for health outcomes). Instrumental variables. Randomized experiments.
33 / 41 Randomized experiments
Setup: Assign individuals/firms/etc randomly into a treatment and control group. E.g. treatment individuals get a better school, control individuals do not. The causal impact is found by comparing outcomes in treatment and control group. Why do we randomize the experiment?
34 / 41 Potential concerns
Not everything can be randomized (e.g. institutions or culture, ethical problems). External validity. General equilibrium effects (e.g. spillovers from treatment to control group). The Hawthorne effect (behavioural response due to treatment).
35 / 41 Tips & tricks : Growth rates
If China grows by 8% for 10 years, what will its GDP be? How long will it take for China to double its GDP?
I To get same income level as Norway? What was China’s average growth rate the last decade?
36 / 41 Growth rates
If x grows at rate g, then
xt+1 = xt (1 + g)
If the growth rate is g in two years, then
xt+2 = xt+1 (1 + g)
= xt (1 + g)(1 + g) 2 = xt (1 + g)
If growth is g in n years, then
n xt+n = xt (1 + g)
37 / 41 Average annual growth
Assume we know xt+n and xt . What is the average annual growth rate? We know n xt+n = xt (1 + g) Rearranging, x 1/n g = t+n − 1 xt
Example: Let xt = 100 and xt+10 = 150. Then
1501/10 g = − 1 = 0.041. 100
38 / 41 n n xNorway (1 + 0.02) = xChina (1 + 0.10) 1 1.02n = 1.10n 6 1 nln1.02 = ln + nln1.10 6 ln(1/6) n = ln1.02 − ln1.10 ≈ 24
An application
China’s GDP/capita is 1/6 of Norway’s GDP/capita. Growth in China is 10% and growth in Norway is 2%. When will China’s GDP/capita surpass Norway’s?
39 / 41 An application
China’s GDP/capita is 1/6 of Norway’s GDP/capita. Growth in China is 10% and growth in Norway is 2%. When will China’s GDP/capita surpass Norway’s?
n n xNorway (1 + 0.02) = xChina (1 + 0.10) 1 1.02n = 1.10n 6 1 nln1.02 = ln + nln1.10 6 ln(1/6) n = ln1.02 − ln1.10 ≈ 24
39 / 41 Continuous time Often convenient to write down models in continuous time. Can be shown that (partial differential equation)
y (t) = y (0)egt
where y (0) is the initial value, g is the growth rate and t is time. Using the properties of logarithms, we get
lny (t) = lny (0)egt = lny (0) + lnegt = lny (0) + gt
−→ The growth rate g is the slope coefficient of e.g. log GDP plotted against time. We denote y˙ = ∂y/∂t and yˆ =y ˙/y. Observe that y˙ = y (0)gegt and yˆ = g.
40 / 41 70-year rule
How long does it take for income to double if growth is constant? We have y (t) = y (0)egt And want to find y (t) = 2y (0), or
y (0)egt = 2y (0) egt = 2 gt = ln2 ln2 0.70 70 t = ≈ = g g % annual growth
If g = 0.02, then t = 35. If g = 0.08, then t ≈ 9.
41 / 41