factors in dampening the downturn in 2014 compared with the more severe decline in output that was experienced in the 2008/2009 global financial crisis. Although these measures have been important steps to deal with the shorter run implications of Russia’s oil dependence, they cannot change the fact that policies aimed at diversifying the economy are the only solutions to generate stable and sustainable growth at a level that is sufficient to close the income gap with high-income countries and stay ahead of its middle-income emerging market peers (Becker, 2018). In its October 2018 forecast of the world economy, the IMF (2018) projected that Russia will grow by 1.8%, similar to the 2% growth in advanced economies but well behind the 4.7% growth in emerging markets.
Are “normal” growth models relevant for Russia?
If we turn our attention to regular growth models that focus on factors that the literature has identified as fundamental drivers of growth, we may better understand what Russia needs to do to boost growth going forward. Becker and Olofsgård (2018) use a robust empirical growth model to understand differences in growth across 25 transition countries in the first 25 years since the dissolution of the Soviet Union. The model was originally specified and estimated by Levine and Renelt (1992) with a focus on identifying the robust determinants of growth among the long list of variables that have been used in empirical growth models. In the end, the authors show that initial GDP, population growth, human capital measured by secondary schooling and the ratio of investments to GDP are the most robust determinants of growth across a large number of countries and over time. The model was estimated without the transition countries that we study in Becker and Olofsgård (2018). We could therefore use the estimated model to see how well it predicted the growth experience of transition countries to investigate the question if (and when) transition countries can be thought of as “normal” countries from a growth perspective.3
Using the same methodology here but with a focus on Russia and the country groups that we used in Figure 1, we can generate predicted growth and compare this with the actual growth for the first decade of transition and then do the same with the 17 years that coincide with Putin being the President and Prime Minister of Russia. For the initial period, the model predicts that Russia would grow at 4.8% per annum, while in fact, income declined by 5% per year on average. Russia thus underperformed the expected growth by almost 10% points per annum. This is similar to the other FSU countries but far behind the EU10 group that “only” underperformed the model by around 5% points.
The picture changes dramatically when we look at the period 2000–2017. Both Russia and the peer groups have growth that comes very close to what the model predicts; the residuals are a few tenths of a percent up or down. In this sense, these countries are in this time period indeed “normal” countries.
The numbers in Table 1 also allow us to discuss the quantitative importance of the different fundamental growth factors in generating the predicted growth rates. The general impression is that human and physical capital as measured by secondary schooling and investments to GDP are of equal importance and of more significance numerically than the other variables. However, the second observation is that there is much less variation in the growth that is generated by human capital than by physical capital. If Russia had the secondary schooling of the average EU10 country, growth would only increase by 0.15% points while if the investment rate was on par with EU10 countries, growth would increase by 1.3% points. In other words, differences in investment to GDP ratios explain almost all of the difference in predicted growth between Russia and the EU10 countries. For the political leaders of Russia, this is an important message. The various proposals to modernise and diversify the economy can have a large impact on expected growth in Russia and with the right incentives to invest in sectors that are less subject to external volatility, this would also make Russia’s growth more robust.4 It is therefore important to understand how investments have evolved over time and how this can be explained. This is the focus in the following sections.
Table 1:Russia and peers — predicted and actual growth.
Source: Becker and Olofsgård (2018) based on Penn World Data 9.0 and additional calculations.
Investments
There are a number of measurement issues related to investments (and other variables) in the national accounts statistics. The data in Table 1 are from the Penn World Table 9.0, where there is an effort to make data comparable between a large number of countries, including Russia. Investments to GDP are measured as the share of gross capital formation at current PPPs and are extremely high in the initial years of transition and much lower later in the sample compared to the official statistics from the Federal Statistics Service. If we use the official data, there are also significant differences in the dynamics of investments between data in current prices or constant prices.
An important factor behind the differences in shares between the current and constant price series is due to the importance of oil exports. The constant price data measure exported quantities, while the current price data measure export values and are therefore subject to changes in both international oil prices (measured in dollars) and changes in the exchange rate (since the accounts are in ruble). Since GDP shares obviously have to add up to 100% (at least when the statistical discrepancy is taken care of), if exports develop very differently for the current and constant price series, so will all the shares, including investments to GDP.
Instead of focusing on how the share of investment in GDP develops, we can look at the growth rate of investment, which is not subject to an adding up constraint. To avoid having inflation that has varied greatly over the years distorting the analysis, growth should be measured in real terms. This implies using either the constant price series or taking the current price series and converting it to dollars with the idea that the exchange rate will move in a direction opposite to that of inflation and provide a measure that is closer to real growth in investments. Since the next step of the analysis involves exploring how capital flows (which are measured in dollars in the balance of payments statistics) are related to investments, the focus will be on how investments measured in dollars have evolved.
The first observation from Figure 4 is that the growth of investments has varied greatly since the start of transition, which is not surprising given the growth charts we have seen. As expected, investment is more volatile than growth but since we are looking at growth in dollar terms, both series display a very high degree of volatility. Although the initial years of transition were particularly volatile with the initial investment boom followed by the 1998 crash, more recent quarters also display growth rates going from plus to minus 40%, which of course is linked to significant changes in the exchange rate.5
Figure 4:Investment and GDP growth.
Note: Growth is calculated from the same quarter last year on GDP measured in current terms and converted to USD by using quarterly exchange rates.
Source: Federal Statistics Service and author’s calculations.
What are then the factors that drive changes in investment? In many transition countries, foreign direct investments have been important drivers of investment and growth (see Mileva, 2008). Russia has of course received large foreign investments since 1991, but in many empirical studies of FDI, Russia
