Tag: long-term growth

Enemies of the People

Posted on May 24, 2021 | Policy Brief

20210524 Enemies of the People FREE Network Image 04

From the early days of the Soviet Union, the regime designated the educated elite as Enemies of the People. They were political opponents and considered a threat to the regime. Between the late 1920s and early 1950s, millions of enemies of the people were rounded up and forcedly resettled to remote locations within the GULAG, a system of labor camps spread across the Soviet Union. In recent research (Toews and Vezina, 2021), we show that these forced relocations have long-term consequences on local economies. Places close to camps that hosted more enemies of the people among prisoners are more prosperous today. We suggest that this result can be explained by the intergenerational transmission of education and a resulting positive effect on local development, which can still be observed to this day.

Historical Background

Targeting the educated elite, collectively referring to them as Enemies of the People and advocating their imprisonment, can be traced back to the beginning of the Russian Revolution in 1917. After consolidating power a decade later, Stalin launched the expansion of the GULAG system, which at its peak consisted of more than a hundred camps with over 1.5 million prisoners (see Figure 1). A large number of historians extensively described this dark episode in Russian history (Applebaum (2012), Khlevniuk (2004), and Solzhenitsyn (1974)). During the darkest hours of this episode, the Great Terror, 1.5 million enemies were arrested in just about two years. While half were executed immediately, the other half were forcedly allocated to GULAG camps spread across the Soviet Union and mixed with non-political prisoners (see Figure 2). Enemies accounted for about a third of GULAG prisoners after the Great Terror. As a result, education levels were higher in the GULAG than in society. In 1939, the share of GULAG prisoners with tertiary education was 1.8%, while, according to the Soviet Census of the same year, only 0.6% of the population had tertiary education.

After Stalin’s death, labor camps started closing rapidly, but many ex-prisoners settled close to the campsites. New cities were created and existing cities in the proximity of camps started growing fast (Mikhailova, 2012). Enemies remained once freed for a combination of political, economic, and psychological reasons. Politically, they were constrained in their choice of location by Stalin-era restrictions on mobility. Economically, they had few outside options and could keep on working for the camps’ industrial projects. On the psychological level, prisoners had become attached to the location of the camp, as Solzhenitsyn (1974) clearly puts it: “Exile relieved us of the need to choose a place of residence for ourselves, and so from troublesome uncertainties and errors. No place would have been right, except that to which they had sent us.”.

Figure 1. Location and size of camps in the Soviet Gulag system

Notes: The circles are proportional to the prisoner population of camps. The data is from the State Archive of the Russian Federation (GARF) and Memorial

Enemies of the People and Local Prosperity

At the heart of our analysis is a dataset on GULAG camps which we collected at the State Archive of the Russian Federation (GARF). It allows us to differentiate between prisoners who were imprisoned for political reasons (Enemies of the People) and those arrested for non-political crimes. The share of enemies varied greatly across camps, and we argue that this variation was quasi-random. We back this up by the historical narrative, according to which the resettlement process was driven by political rather than economic forces, suggesting that strategic placements played little role in the allocation of enemies (Khlevniuk (1995) and Ertz (2008)). Moreover, while the forced nature of allocation to camps allows us to rule out endogenous location decisions, we also show that neither economic activities nor geographic attributes, such as climatic conditions, soil quality, or the availability of resources, predict the share of enemies across camps.

To estimate the long-run effects of enemies on local prosperity, we link the location of camps in 1952, the year before Stalin’s death and at the peak of the GULAG system, to post-Soviet data covering the period 2000-2018.

Figure 2. The rise and fall of the Gulag

Notes: The solid line shows the number of Gulag camps while the dashed line shows the total number of prisoners in the Gulag. The two vertical dash lines indicate the years that can define the start and end of the Gulag, starting with Stalin’s 5-year plan in 1928 and ending with Stalin’s death in 1953. The shaded areas show specific periods of marked change for the Gulag, starting with dekulakization in 1929, when when Stalin announced the liquidation of the kulaks as a class and 1.8 million well-off peasants were relocated or executed. The Great Terror of 1936-1938, also referred to as the Great Purge, is the most brutal episode under Stalin’s rule, when 1.5 million enemies were arrested, and half of them executed. The Gulag’s prisoner population went down during WW2, as non-political prisoners were enlisted in the Red Army, and as the conditions in camps deteriorated and mortality increased. Source: Memorial.

In particular, the camp level information is linked to the location of firms from the Russian firm census (2018), data on night-lights (2000-2015), as well as data from household and firm-level surveys (2016 and 2011-2014, respectively). Our results suggest that one standard deviation (28 percentage point) increase in the share of enemies of the people increases night-lights intensity per capita by 58%, profits per employee by 65%, and average wages by 22%. A large number of specifications confirm the relationship depicted in Figure 3, which illustrates the positive association between the share of enemies across camps and night-lights intensity per capita.

Figure 3. Share of enemies vs. night lights per capita across Gulags

Notes: The scatters show the relationship between the share of enemies in camps in 1952 and night lights per capita within 30 km of camps in 2000 and 2015. Each circle is a 30km-radius area around a camp, and the size of the circles is proportional to the camps’ prisoner populations. The biggest circle is Khabarovsk. The solid lines show the linear fit, and the shaded areas show the 95% confidence interval. Areas near camps with a higher share of enemies have brighter night lights per capita both in 2000 and 2015. The data on Gulags is from the State Archive of the Russian Federation (GARF) and the data on night lights is from the DMSP-OLS satellite program and made available by the Earth Observation Group and the NOAA National Geophysical Data Center. The data on population is from the gridded population of the world from SEDAC.

Intergenerational Transmission

We suggest that the relationship between enemies and modern prosperity is due to the long-run persistence of high education levels, notably via intergenerational transmission, and their role in increasing firm productivity. For the identification of the intergenerational link, we rely on a household survey collected by the EBRD in which interviewees are explicitly asked whether their grandparents have been imprisoned for political reasons during Soviet times. Exploiting this information, we show that the grandchildren of enemies of the people are today relatively more educated. We also find that grandchildren of enemies are more likely to be residing near camps that had a higher share of enemies of the people among prisoners in 1952. An alternative explanation for our results could be that the leadership of the Soviet Union may have strategically chosen to invest more during the post-GULAG period in locations that had received more enemies to exploit complementarities between human and physical capital. We find no evidence for this mechanism. We document that Soviet investment in railroads, factories of the defence industry, or universities was, if anything, lower in places with a large share of enemies.

Conclusion

We show that the massive and forced re-allocation of human capital that took place under Stalin had long-run effects on local development. Sixty years after the death of Stalin and the demise of the GULAG, areas around camps that had a higher share of enemies are richer today, as captured by firms’ wages and profits, as well as by night-lights per capita. We argue that the education transferred from forcedly displaced enemies of the people to their children and grandchildren partly explains variation in prosperity across localities of Russia. This can be seen as a historical natural experiment that identifies the long-run persistence of higher education and its effect on long-run prosperity. Sadly, it also highlights how atrocious acts by powerful individuals can shape the development path of localities over many generations.

Bibliography

  • Applebaum, A., Gulag: A History of the Soviet Camps, Penguin Books Limited, 2012.
  • Ertz, Simon. Making Sense of the Gulag: Analyzing and Interpreting the Function of the Stalinist Camp System. No. 50. PERSA Working Paper, 2008.
  • Khlevnyuk, Oleg, “The objectives of the Great Terror, 1937–1938.” In Soviet History, 1917–53, pp. 158-176. Palgrave Macmillan, London, 1995.
  • Khlevnyuk, Oleg, The History of the Gulag: From Collectivization to the Great Terror Annals of Communism, Yale University Press, 2004.
  • Mikhailova, Tatiana, “Gulag, WWII and the long-run patterns of Soviet city growth,” 2012.
  • Solzhenitsyn, Aleksandr, The Gulag Archipelago, 1918-56: An Experiment in Literary Investigation, New York: Harper Row, 1973.
  • Toews, Gerhard, and Pierre-Louis Vézina. “Enemies of the people.” (2021).

Disclaimer: Opinions expressed in policy briefs and other publications are those of the authors; they do not necessarily reflect those of the FREE Network and its research institutes.

Russia and Oil — Out of Control

Posted on October 31, 2016 | Policy Brief

Free Policy Brief Image - Russia and Oil — Out of Control

Russia’s dependence on oil and other natural resources is well known, but what does it actually mean for policy makers’ ability to control the economic fate of the country? This brief provides a more precise analysis of the depth of Russia’s oil dependence. This is based on a careful statistical analysis of the immediate correlation between international oil prices — that Russia does not control — and Russian GDP, which policy makers would like to control. I then look at how IMF’s forecast errors in oil prices spillover to forecast errors of Russian GDP. These numerical exercises are striking; over the last 25 years oil price changes explain on average two thirds of the variation in Russian GDP growth and in the last 15 years up to 80 percent of the one-year ahead forecast errors. Instead of controlling the economic fate of the country, the best policy makers can hope for is to dampen the short-run impact of oil price shocks. A flexible exchange rate and fiscal reserves are key volatility dampers, but not sufficient to protect long-term growth. The latter will always require serious structural reforms and the question is what needs to happen for policy makers to take action to get control over the long-term fate of the economy.

In a recent working paper (Becker, 2016), I take a careful look at the statistical relationship between Russian GDP and international oil prices. This brief summarizes this analysis and its policy conclusions.

Russia and oil, the basics

Although Russia’s oil dependence is discussed every time international oil prices drop, it is not uncommon to hear that oil is not really so important for the Russian economy. The argument is that the oil and natural resource sector only accounts for some 10 percent of Russian production. This is indeed consistent with the official sectoral breakdown of GDP that is shown in Figure 1 where the minerals sector indeed only has a 10 percent share.

Figure 1. Structure of GDP in 2015

slide1Source: Federal State Statistics Service, 2016

However, this static picture of production shares does not translate into a dynamic macro economic model that allows us to understand what is driving Russian growth. Instead a careful analysis of the time series of Russian GDP is required to understand how important oil is for growth.

Russian GDP can be measured in many different ways: nominal rubles, real rubles, U.S. dollars, or in purchasing power parity (PPP) terms to mention the most common. Here we focus on GDP measured in real rubles and U.S. dollars since we want to get rid of Russian inflation, which has been quite high for most of the studied time period. The PPP measure generates figures and numerical estimates that are in between the real ruble and U.S. dollar measures and are not included here to conserve space.

The first evidence of the importance of international oil prices as a major determinant of Russian income at the macro level is presented in Figures 2 and 3 where the first figure shows dollar income and the second real ruble income. In both cases it is obvious that there is a strong correlation and that the correlation is higher for income measured in dollars.

Figure 2. U.S. dollar GDP and the oil price

slide2Source: IMF, 2016

Figure 3. Real ruble GDP and the oil price

slide3Source: IMF, 2016

However, it is also clear that all the time series have some type of trends or in econometric language, are non-stationary. This means that simple correlations of the time series shown in Figure 2 and 3 may not be statistically valid (or “spurious” as it is called in the literature). This is not a critical issue but can be handled by regular econometric methods.

Russia and oil, the econometrics

When time series are non-stationary they need to be transformed to some stationary form before we can do regular regressions (in Becker, 2016 I also address the issue of using a framework that allows for co-integration).

Two transformations that make the variables stationary are to use first differences or percent growth rates. Both are used before we run simple regressions of growth or first differences of GDP on growth or first difference in international oil prices. The full sample starts in 1993, but since the early years of transition were subject to many different shocks at the same time, a shorter sample starting in 2000 is also used.

A number of observations come from the estimates that are presented in Table 1: Oil prices are always statistically significant; the adjusted R-squared is higher for dollar income than real rubles (with one exception due to a large outlier in 1993); overall the explanatory power of these simple regressions are very high (42-92 percent) and the explanatory power increases in all specifications when going from the full sample (1993-2015) to the more recent sample (2000-2015). Note that the latter sample perfectly overlaps with the current political leadership so contrary to some wishes; the oil dependence has not been reduced under Putin/Medvedev.

Table 1. Russian macro “models”

slide4Source: Becker 2016

Russia and oil, the forecasts

The strong correlation between international oil prices and Russian GDP provides a very simple econometric model for explaining past variations in Russian GDP. Unfortunately it does not imply that it is easy to forecast Russian GDP since international oil prices are very hard to predict. There are many models that have been used to forecast oil prices, but the IMF and many others now use the market for oil futures to generate its central forecast of oil prices.

The IMF also provides confidence intervals around the central forecast, and the uncertainty surrounding the forecast is substantial: In the latest forecast the 68 percent confidence interval goes from around 20 dollars per barrel to 60 one year ahead, while the 98 percent interval ranges from 10 dollar per barrel to around 85. With oil currently around 45 dollars per barrel, these variations imply that oil prices could either halve or double in the next year, not a very precise prediction to base economic policy on for Russia since the estimates for real ruble growth in the later sample in Table 1 imply that Russian GDP growth in real ruble terms could be anywhere from minus 5 to plus 10 percent, or a fifteen percentage point difference!

If we look at past IMF forecasts of oil prices and Russian GDP and see how much they deviate from actual values a year later we can compute one year ahead forecast errors. We can do this calculation for the last 16 years for which the IMF data is available. Figures 4 and 5 show how the forecast errors in oil prices correlate with the forecast errors for dollar income and real ruble income, respectively. Similar to the regressions presented in Table 1, the correlations are very high for both measures of GDP: 82 percent for dollar GDP, and 65 percent for real ruble GDP.

In other words, a very large share of the uncertainty surrounding Russian GDP forecasts can be directly attributed to variations in international oil prices, a variable that (again) Russia does not control. The fact that the variations in oil prices explain somewhat more of the variation in dollar income compared to real ruble income is a result of a policy change that in later years allowed the exchange rate to depreciate much more rapidly when oil prices fall.

Figure 4. Forecast errors

slide5Source: Becker 2016

Figure 5. Forecast errors

slide6Source: Becker 2016

Policy conclusions

The depth of Russia’s oil dependence is much greater than what casual observers of the mineral sectors share of GDP would suggest. At the macro level, variations in international oil prices explain at least two thirds of actual Russian growth and even more of the one-year ahead forecasts errors.

The experience of the 2008/09 global financial crisis provided an important lesson to Russian policy makers, which is that exchange rate flexibility is required to dampen the real impact of falling oil prices and to protect both international reserves and the fiscal position. In the more recent years, the currency has been allowed to depreciate in tandem with falling oil prices and the drop in real ruble income was therefore less severe in 2015 than in 2009. Income in dollar terms, instead, took a greater hit, but this was a necessary corollary to protecting reserves and the budget. A flexible exchange rate and gradual move to inflation targeting in combination with accumulating fiscal reserves in times of high oil prices are key to Russia’s macro economic stability.

Nevertheless, these policies are not sufficient to remove the long-run impact that low or declining oil prices will have on growth, measured both in real ruble terms or dollar terms. It is nice to have fire insurance when your house burns down, but when you rebuild the house you may want to consider not building another straw house. For Russia to build a strong economy that is not completely hostage to variations in international oil prices, fundamental reforms that encourage the development of alternative, internationally competitive, companies are needed. This includes reforms that initially will reduce policy makers control over the economy and legal system, but over time it will provide the much needed diversification away from exporting oil that puts the fate of the Russian economy squarely in the hands of international oil traders. Losing some control today may provide a lot more control in the future for the country as a whole, but perhaps at the expense of less control for the ruling elite.

References