Global job creation remains sluggish, prompting calls for policy actions to raise economic growth. Will growth create jobs? Recent IMF research documents a striking variation among countries in the extent to which employment responds to GDP growth over the course of a year. In some countries, labor markets are quite responsive: when growth picks up, employment goes up and unemployment falls; in other countries the response is quite muted. Thus, a pick-up in growth—through aggregate demand stimulus for instance—will result in more jobs, but the extent of job creation in the short run could vary sharply across countries. Some structural measures can thus serve as useful complementary policies, as also discussed in IMF research.
The global unemployment rate increased sharply during the Great Recession to a peak of 6.2 percent in 2009 (Figure 1a). The increase in unemployment in emerging markets and developing countries was modest. In contrast, the unemployment rate in advanced countries shot up to nearly 8.5 percent and has returned very sluggishly in subsequent years toward its pre-crisis level. Even by 2017, the forecast is that unemployment in advanced economies will still be higher than it was a decade earlier (Figure 1b).
Figure 1a:Global Unemployment Rates Figure 1b:Global Unemployment Rates
Notes: The vertical scale shows unemployment rates in percent. Figure 1a shows the global unemployment rate, Figure 1b the rates for sub-groups: advanced, emerging markets and low-income developing countries (LIDC).
Source: International Jobs Report, April 2016, OCP Policy Center.
Some observers argue that the sluggish recovery of unemployment in advanced economies is to be expected given that output remains below trend in many of these countries; they argue further that the way to reduce unemployment faster would be through macroeconomic policies (e.g., Krugman 2011, Kocherlakota 2014). Others suggest that the job losses are not well explained by developments in output (e.g., McKinsey 2011, Cazes and others 2011). Recent IMF research provides evidence on this debate by documenting the short-run relationship between labor market developments and real GDP movements for a large group of countries. Such a relationship is often referred to as Okun’s Law, following Okun’s (1962) documentation of it for the United States.
Jobs and Growth in Advanced Economies
Figure 2a shows the relationship between unemployment changes and GDP growth for the United States. The relationship documented by Okun holds up very well even with the addition of over 50 years of data, cementing its claim to be called a “Law.” The relationship also holds well for most U.S. states, with the strength of the relationship dependent in part on the state’s industrial structure. In the so-called Rust Belt states, unemployment is very responsive to cyclical fluctuations in the economy, but the relationship is weaker in states where agriculture or oil production are dominant (Figure 2b).
Figure 2a:Does Growth Lower Unemployment? U.S. Evidence Figure 2b:Does Growth Lower Unemployment? U.S. Evidence
Notes: In Figure 2a, the horizontal axis shows real GDP growth in percent; the vertical axis shows the change in the unemployment rate in percentage points. Figure 2b shows the change in unemployment in response to real GDP growth in various groups of U.S. states.
While Okun’s Law holds well overall for the United States, the behavior of unemployment since 2011 did deviate from the historical relationship because of an unusual fall in labor force participation (Schindler and others 2014). Erceg and Levin (2013) emphasize the unusual depth and duration of the Great Recession as a reason: in their view, participation normally does not respond much to output fluctuations given the costs of entering and exiting the labor force, but a protracted recession eventually leads workers to exit.
Another reason for the sluggish decline in U.S. unemployment is adverse shocks to particular sectors, particularly construction and finance. There is evidence that these shocks, and the uncertainty generated about sectoral fortunes, were particularly important in accounting for long-duration unemployment (see Chehal, Loungani, and Trehan 2010; IMF 2010; and Choi and Loungani 2015). Though the U.S. unemployment rate has now dipped below 5 percent, there are still some structural issues in the labor market. Mobility across states in response to adverse shocks has been declining—and indeed may not have been as high in the first place as suggested in previous studies (Dao, Furceri, and Loungani 2016).
Cyclical unemployment in most European economies can also be explained well by Okun’s Law (Arpaia, Kiss, and Turrini 2014; and Bakker 2016), as can a large part of the increase in youth unemployment seen during the Great Recession (Banerji, Saksonovs, Lin, and Blavy 2014).
Evidence for Emerging Markets and Low-Income Countries
When looking across a broad group of economies, employment may be a better labor market indicator than unemployment. It is more likely to be measured in a comparable way across countries and the data may be of better quality; moreover, in low-income countries, unemployment may not be an option for many people. Figure 3a shows the so-called Okun coefficient—how much employment increases when growth picks—for the G20 economies, which together account for the lion’s share of global GDP and employment.
Figure 3a:Does Growth Create Jobs? Evidence for Country Groups Figure 3b:Does Growth Create Jobs? Evidence for Country Groups
Notes: Both panels show the response of employment growth to real GDP growth. India is not shown in Figure 3a due to lack of data. In Figure 3b, LICs includes countries classified by the World Bank as low-income and lower-middle income countries.
In South Africa, Australia, and Canada, a 1 percent increase in GDP is matched by an increase in employment of 0.6 percent or higher. In contrast, there is virtually no response of employment to growth in China, Indonesia, and Turkey. The extent to which changes in growth account for changes in unemployment and employment also varies across countries. GDP growth accounts for over 70 percent of the variation in employment in Canada and the United States, about 40 percent in Russia, the United Kingdom, and Australia, and very little in many other countries.
Figure 3b shows the average value of the Okun coefficient for three groups of countries: advanced, emerging, and low-income (LICs). In addition to the average, the range of values for different countries within the group is shown. Among advanced economies on average, employment increases by 0.4 percent for a 1 percent in GDP growth and the variation across countries is small. In emerging markets, the average value is smaller, 0.2 percent, though again with some variation across countries. For LICs, the average coefficient is small—barely above zero—and with a very large variation across countries (Figure 4).
Figure 4:Does Growth Create Jobs: Evidence for Emerging Markets and LICs
Note: The bars show the response of employment growth to real GDP growth.
Summing up the Evidence
There is often a maintained assumption that labor market outcomes in many countries are largely determined by longer-term structural factors than by short-term cyclical fluctuations. Our results give an initial diagnostic check on the validity of this view. There are several countries in which Okun’s Law holds poorly in the sense that the estimate of the Okun coefficient is low in absolute value and the overall fit of the equation is poor. In these countries the assumption of a dichotomy between output and labor market fluctuations may be fine as a starting point. But in any more cases, where Okun’s Law holds well, the view is less tenable. At the IMF, a
To sum up, for the majority of countries around the globe, taking account of growth is an important part of understanding short-run unemployment fluctuations. In the case of other countries, there are several possible explanations for the weakness of the jobs-growth link. In some cases, reported unemployment rates may not fully reflect the true unemployment rate. Some countries are going through rapid structural change and unemployment may be driven by this longer-run trend rather than short-run fluctuations. For instance, this is likely to be the case in Morocco, where the unemployment rate has fallen sharply over the past 20 years with the increase in trend GDP but the short-run responsiveness of unemployment to GDP growth is essentially zero. In countries with large rural sectors and a large degree of informality, the measured unemployment rate (which is more likely to reflect urban and formal sectors) may not be very responsive to growth.
Policy Implications: A Two-Handed Approach
The evidence that extra growth will bring back jobs in many countries leads to the obvious question: what will deliver the extra growth? In Furceri and Loungani (2014), we advocate a two-handed approach: continued support to domestic aggregate demand and the adoption of policies and reforms that can boost aggregate supply. Without supportive demand policies, supply measures could have little impact in the short run. If companies do not see improved sales prospects, they will not increase capacity; hence, it is essential to ensure that the demand is there to sustain supply. But without supply measures, output gains based solely on a stimulus to demand will prove temporary. The range of supply measures varies, from removing bottlenecks in the power sector to reforms in labor and product markets (IMF 2016; Adhikari, Duval, Hu, and Loungani 2016). In many countries, there is a strong case for increasing public infrastructure investment, which would provide a much-needed boost to demand in the short term and would also help supply (i.e., potential output) over the longer term (Abiad and others 2014).
AdhikariBibekRomainDuvalBingjieHu and PrakashLoungani. 2016. “Can Reform Waves Turn the Tide? Some Case Studies Using the Synthetic Control Method,” IMF Working Paper (forthcoming).
ArpaiaAlfonsoAronKiss and AlessandroTurrini. 2014. “Is Unemployment Structural or Cyclical?: Main Features of Job Matching in the EU After the Crisis.” Economic Papers527: 1–60.
BakkerBas. 2016. “Employment and the Great Recession: The Role of Real Wages,” IMF Working Paper 15/229International Monetary FundWashington DC.
IBallLaurenceDavideFurceriDanielLeigh and PrakashLoungani. 2016. “Does One Law Fit All? Cross-Country Evidence on Okun’s LawIMF Working Paper (forthcoming).
BallLaurenceJoãoTovar Jalles and PrakashLoungani. 2015. “Do Forecasters Believe in Okun’s Law? An Assessment of Unemployment and Output Forecasts.” International Journal of Forecasting31(1): 176–184.
BallLaurence M.DanielLeigh and PrakashLoungani. 2013. Okun’s Law: Fit at Fifty?. Working Paper No. w18668. National Bureau of Economic ResearchCambridge, MA.
BanerjiAnganaSergejsSaksonovsHannahHuidan Lin and RodolpheBlavy. 2014. Youth Unemployment in Advanced Economies in Europe: Searching for Solutions. SDN 14/11. International Monetary FundWashington, DC.
CazesSandrineSherVerick and FaresAl-Hussami. 2011. “Diverging Trends in Unemployment in the United States and Europe: Evidence from Okun’s Law and the Global Financial Crisis.” Employment Working PapersILO.
ChamiRalphYasserAbdihAlbertoBeharSerhanCevikLisaDougherty-ChouxDavideFurceriNickJanus and PaulZimand. 2012. “A Template for Analyzing and Projecting Labor Markets” TNM 12/01. International Monetary FundWashington, DC.
ChehalPuneetPrakashLoungani and BharatTrehan. 2010 “Stock-Market-Based Measures of Sectoral Shocks and the Unemployment Rate.” FRBSF Economic Letter (2010): 23.
ChoiSangyup and PrakashLoungani. 2015. “Uncertainty and Unemployment: The Effects of Aggregate and Sectoral Channels.” Journal of Macroeconomics46: 344–358.
DaoMaiDavideFurceri and PrakashLoungani. 2014. Regional Labor Market Adjustments in the United States. IMF Working Paper 14/26. International Monetary FundWashington DC.
ErcegChristopher J. and LevinAndrew T.2013. “Labor Force Participation and Monetary Policy in the Wake of the Great Recession” IMF Working Paper 13/245. International Monetary FundWashington, DC.
FurceriDavide and PrakashLoungani. 2014. “Growth: An Essential Part of the Cure for Unemployment” iMFDirect Blog. (November192014).
Gonzalez PrietoNathaliePrakashLoungani and SaurabhMishra. 2016 “What Lies Beneath: A Sub-National Look at Okun’s Law for the United States.” IMF Working Paper (forthcoming).
IMF. 2010. “Chapter 3. Unemployment Dynamics During Rrecessions and Recoveries: Okun’s Law and Beyond.” In World Economic Outlook: Rebalancing Growth.Washington DCApril.
IMF. 2014. “Chapter 3. Is It Time for an Infrastructure Push? The Macroeconomic Effects of Public Investment.” World Economic Outlook: Legacies Clouds Uncertainties.Washington DCOctober.
IMF. 2016. “Chapter 3. Time for a Supply-Side Boost? Macroeconomic Effects of Labor and Product Market Reforms in Advanced Economies.” World Economic Outlook: Too Slow for Too LongApril.
KocherlakotaNarayana. 2014. Opening Remarks, presented at “Conversations with the Fed” on January92014. Federal Reserve Board of Indianapolis.
KrugmanPaul. 2011. “The Fatalist Temptation,” New York TimesJuly9.
LounganiPrakash and SaurabhMishra. 2015. Does Growth Create Jobs in the G-20 Economies?. Policy Brief No. 1522.OCP Policy CenterRabat, Morocco.
McKinsey Global Institute. 2011.) “An Economy That Works: Job Creation and America’s Future,” June2011.
OCP Policy Center. 2016. International Jobs Report: May 2016. No. 1602.
OkunArthur M.1962. “Potential GNP: Its Measurement and Significance.” Reprinted as Cowles Foundation Paper 190. Yale UniversityNew Haven.