Abstract

In the baseline, supportive monetary policy and medium-term fiscal consolidation remain valid for many economies in the region. In the event of a negative growth shock, monetary policy should be the first line of defense, while automatic fiscal stabilizers should be allowed to play freely, provided there is enough fiscal policy room to do so. In case of a major shock and depending on the nature of the shock, fiscal policy should ease within the medium-term adjustment plans that dispel concerns about sustainability. Against the backdrop of mediocre global growth prospects, structural reforms are critical to lift potential growth and re-accelerate convergence.

In the baseline, supportive monetary policy and medium-term fiscal consolidation remain valid for many economies in the region. In the event of a negative growth shock, monetary policy should be the first line of defense, while automatic fiscal stabilizers should be allowed to play freely, provided there is enough fiscal policy room to do so. In case of a major shock and depending on the nature of the shock, fiscal policy should ease within the medium-term adjustment plans that dispel concerns about sustainability. Against the backdrop of mediocre global growth prospects, structural reforms are critical to lift potential growth and re-accelerate convergence.

In the absence of negative shocks, supportive monetary policy and medium-term fiscal consolidation is an appropriate policy mix for many CESEE economies:

  • Monetary policy should stay accommodative in low-inflation countries with further rate cuts if inflation expectations continue to decline or interest rate differentials with the euro area widen. There are, however, many countries in the region, where conventional monetary policy space is limited: because they lack monetary policy autonomy or because inflation is above the target or the interest rate at zero bound. For Turkey, where inflation remains high, further tightening would be needed to address excess demand pressures and build up international reserves that are now below the IMF’s metric of reserve adequacy. In Russia, a resumption of monetary easing to support the weak economy is only feasible once inflation expectations fall. The pace of such future easing would need to be mindful of the uncertain external outlook and the need to build the credibility of the new inflation-targeting regime. About half of the countries in the region that have some sort of flexible exchange rate regime can also use this tool to counter adverse external shocks. However, high foreign currency indebtedness, euroization and financial openness could limit the benefits of currency depreciations.

  • Fiscal policy should continue to anchor medium-term debt sustainability and build policy buffers in most countries. Some have made progress with a noticeable decline in the structural fiscal deficit. However, with the debt-stabilizing primary gap still in the negative for several countries and public debt still high, more needs to be done to rebuild fiscal buffers in the medium term (Figure 3.1). For Russia, the roughly neutral fiscal stance remains appropriate, given cyclical weakness, but more consolidation is needed over the medium-run. In Turkey, a tightening of fiscal policy in the medium term would increase domestic savings and thereby soften excess demand pressures, while building more policy space.

  • Medium-term fiscal consolidation should rely, as much as possible, on more growth-friendly expenditure and revenue measures, as discussed in the Fall 2015 REI. On the expenditure side, it is important to reduce unproductive transfers and further reform entitlement programs, including public pension systems, while protecting productive spending on public investment. Restructuring of public employment may also be called for, especially where employment levels or public sector wages are higher than in the private sector. On the revenue side, policymakers should consider the introduction or strengthening of carbon and property taxes, and in some cases, the improvement of tax compliance and administration.

Figure 3.1.
Figure 3.1.

Estimated Remaining Adjustment Needs

(Percent of GDP)

Sources: World Economic Outlook, and IMF staff calculations and projections.Note: The remaining adjustment needs reflect values for primary balance and structural balance as of end-2015 (negative values represent no adjustment need based on that particular measure). For Ukraine data refers to 2016. -1 percent of GDP is European Commission’s Medium Term Objective (MTO) for many but not all CESEE countries and actual adjustment needs based on country-specific MTO may be different. Debt-stabilizing primary balance is the ratio of primary balance to GDP that stabilizes the debt to GDP ratio at its projected 2021 value.

If growth and inflation surprise on the downside, monetary policy should be the first line of defense. Also, automatic fiscal stabilizers should generally be allowed to operate freely. In case of a very adverse external demand shock, fiscal stimulus may need to be deployed by countries that still have access to international capital markets on affordable terms. For those with this option, it is recommended that they rely on measures that are easy to pull back if economic conditions improve (for example, a temporary investment tax credit) or that enhance economy’s long-term growth potential (for example, targeting infrastructure). More generally, for such stimulus to be effective and not raise questions about sustainability, it should be overlaid on medium-term adjustment plans that noticeably reduce public debt. Deploying the latter, together with further repair of balance sheets – as discussed in the Spring 2015 REI, is the main macroeconomic policy challenge for many CESEE economies.

Against the backdrop of mediocre global growth prospects, structural reforms are critical to lift potential growth and re-accelerate convergence. As discussed in Chapter II, efforts should focus on active labor market policies and productivity-enhancing reforms. The analysis in this report suggests that the currently significant productivity gaps with advanced Europe could be reduced by upgrading institutions (protection of property rights, legal systems, healthcare), increasing the affordability of financial services (for small and productive firms), and improving government efficiency. While the structural reform recommendations in IMF country reports are generally more comprehensive9 and more tailored to country-specific circumstances, many of the cross-country themes are similar to the ones highlighted in this report (Figure 3.2): improving government efficiency and reducing regulatory burden on firms in most CESEE countries; strengthening governance and institutions in SEE and the CIS; and increasing labor force participation in CEE countries, while improving labor market flexibility in SEE countries. As discussed in the WEO, in cases where the necessary reforms could have negative short-term impact on growth in the context of significant economic slack, these negative effects would need to be mitigated through careful phasing or demand support, if possible10.

Figure 3.2.
Figure 3.2.

IMF Country Teams' Recommendations on Structural Reform Priorities

Source: Latest IMF Country Reports.Note: CEE = Central and Eastern Europe; CESEE = Central, Eastern, and Southeastern Europe; CIS = Commonwealth of Independent States; SEE = Southeastern Europe; SEE-XEU = Southeastern European countries outside the EU.

In countries with greater structural challenges more far-reaching reforms may be needed to speed up convergence. As discussed in IMF country reports, reforms in SEE non-EU and CIS economies should aim to strengthen governance, to lower administrative and trade barriers, increase competition in domestic markets, and improve the transparency and efficiency of public investment procedures. In Belarus, deep and carefully sequenced structural reforms to re-orient the economy toward more private-sector-led growth remain critical. For Moldova, priority should be given to strengthening institutional quality. For Ukraine, critical reforms include anti-corruption and judicial measures, tax administration reforms, and reforms of state-owned enterprises to improve corporate governance and reduce fiscal risks.

Annex I. CESEE: Growth of Real GDP, Domestic Demand, Exports, and Private Consumption

(Percent)

article image
Source: IMF, World Economic Outlook database, Spring 2016 published version.

Weighted averages using 2014 GDP valued at purchasing power parity.

Includes Albania, Belarus, Bosnia and Herzegovina, Bulgaria, Croatia, Czech Republic, Estonia, Hungary, Kosovo, Latvia, Lithuania, Macedonia FYR, Moldova, Montenegro, Poland, Romania, Russia, Serbia, Slovak Republic, Slovenia, Turkey, and Ukraine.

CESEE excluding Czech Republic, Estonia, Latvia, Lithuania, Slovak Republic, and Slovenia.

Includes Bulgaria, Croatia, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovak Republic, and Slovenia.

Annex II. CESEE: Consumer Price Index Inflation, Current Account Balance, and External Debt

(Percent)

article image
Source: IMF, World Economic Outlook database, Spring 2016 published version.

Weighted averages using 2015 GDP valued at purchasing power parity.

Includes Albania, Belarus, Bosnia and Herzegovina, Bulgaria, Croatia, Czech Republic, Estonia, Hungary, Kosovo, Latvia, Lithuania, Macedonia FYR, Moldova, Montenegro, Poland, Romania, Russia, Serbia, Slovak Republic, Slovenia, Turkey, and Ukraine.

CESEE excluding Czech Republic, Estonia, Latvia, Lithuania, Slovak Republic, and Slovenia.

Includes Bulgaria, Croatia, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovak Republic, and Slovenia.

Annex III. CESEE: Evolution of Public Debt and General Government Balance

(Percent of GDP)

article image
Source: IMF, World Economic Outlook database, Spring 2016 published version.

As in the WEO, general government balances reflect IMF staff’s projections of a plausible baseline, and as such contain a mixture of unchanged policies and efforts under programs, convergence plans, and medium-term budget frameworks. General government overall balance where available; general government net lending/borrowing elsewhere. Public debt is general government gross debt.

Weighted averages using 2015 GDP valued at purchasing power parity.

Reported on a cash basis.

Regarding the overall balance, this includes fiscal room for donor-financed capital projects (for 2016-2018 period), which might not be fully utilized by year-end. Public debt includes former Yougoslav debt, not yet recognized by Kosovo.

General government balance: the measure reflects augmented balance, which adds to the balance of general government outlays for banks recapitalizations and is related to called guarantees of publicly-guaranteed debt.

Includes Albania, Belarus, Bosnia and Herzegovina, Bulgaria, Croatia, Czech Republic, Estonia, Hungary, Kosovo, Latvia, Lithuania, Macedonia FYR, Moldova, Montenegro, Poland, Romania, Russia, Serbia, Slovak Republic, Slovenia, Turkey, and

CESEE excluding Czech Republic, Estonia, Latvia, Lithuania, Slovak Republic, and Slovenia.

Includes Bulgaria, Croatia, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland, Romania, Slovak Republic, and

Annex IV. The Determinants of Saving Rates in CESEE EU Countries11

Table 1.

Comparison of Household Saving Rate Determinants in CESEE and Advanced Europe

article image
Sources: Eurostat; World Bank; and IMF staff calculations.Note: CESEE = Central, Eastern, and Southeastern Europe. High-saving CESEE countries (>5 percent) are Croatia, Czech Republic, Hungary, Poland, Slovak Republic, and Slovenia. Medium-saving CESEE countries (>0 percent) are Estonia Latvia, and Lithuania. Low-saving CESEE countries (<0 percent) are Bulgaria and Romania.

Expected sign based on literature review.

Correlations between saving rates and their potential determinants in the sample of CESEE and AE EU countries.

Table 2.

Comparison of Corporate Saving Rate Determinants in CESEE and Advanced Europe

article image
Sources: IMF, Eurostat, World Bank. IMF staff calculations.Note: High-saving CESEE countries (>30 percent) are Croatia, Czech Republic, Estonia, Hungary, Lithuania, and Slovenia. Medium-saving CESEE countries (>20 percent) are Bulgaria, Latvia, Poland, Romania and the Slovak Republic. FDI = foreign direct investment.

Expected sign based on literature review.

Correlations between saving rates and their potential determinants in the sample of CESEE and advanced European Union countries.

Annex V. Derivation of the Golden Rule Benchmark12

In the neoclassical (Ramsey-Cass-Koopmans) growth model, an economy converges to its steady-state equilibrium in which consumption is maximized, the saving/investment rate is constant at its “golden-rule” value, and income, consumption, and capital all grow at a fixed rate equal to the sum of the exogenous growth rates of the labor force and labor-augmenting productivity.

In the Ramsey-Cass-Koopmans model, the “golden-rule” of capital accumulation is given by:

SY=IY=α(δ+n+g)(p+δ+n+g)

where α is the capital share of output; p is the social rate of time preference; δ is the depreciation rate; n is the growth of the labor force; and g is the rate of technical progress.

The model is calibrated for European Union (EU) countries using national accounts data, labor market surveys, and the Penn World Tables Version 8.1 (PWT). In the case of EU members, for which standardized data exist from Eurostat, the estimates of the growth of the labor force and the capital share of output are derived from labor market surveys and national accounts data, respectively. Similar to the adjustments made to the raw data in the PWT, the estimates of the labor share of output are augmented by 63 percent of self-employment income. For non-EU members, data for these variables are from the PWT. In addition, estimates of capital stocks, depreciation rates, and total factor productivity are taken from the PWT.

Under a typical calibration and with a starting value of the capital-to-labor ratio below the steady state, the model implies that the investment rate would fall monotonically toward the “golden-rule” as the economy converges to its steady state. As such, the closed-economy, golden-rule saving/investment can be interpreted as a lower bound for the investment rate along its path of convergence to euro area income levels. The social rate of time preference is constant and set equal to 5 percent for all CESEE countries. This value corresponds to the social rate of time preference in the euro area, derived from the golden rule under the assumption that the euro area has been close to its steady-state path of development on average over 2002–14.

Annex VI. Derivation of the Historical Benchmark13

The purpose of the benchmark is to provide a proxy for a sustainable path of the investment rate during the transition to a steady state. Although neoclassical growth theory does not offer a closed-form solution for such transition dynamics, the “catch-up” is essentially driven by differences in real interest rates that affect intertemporal choices of consumption and savings (the Euler equation; see Barro and Sala-i-Martin 2003). When relative capital scarcity makes capital more productive, bearing a higher real interest rate, it stimulates saving and investment rates and leads to faster pace of capital accumulation. With a rising K/L ratio, the real return to capital declines and saving and investment rates gradually fall to their steady-state constant level. The further the economy is from its steady-state K/Y ratio, the faster it will accumulate capital. Therefore, the transition path for the investment rate I/Y may be approximated by a function of the real return to capital (given by the marginal product of capital, using Cobb-Douglas production function, where A is labor-augmenting productivity, K is capital, L is labor, and α is the capital share):
ItYtf(MPK)=α(KtAtLt)α1ln(ItYt)=c+βln(At)(1α)ln(KtLt)+εt
and where in the steady-state c equals ln(α) and β equals (1 − α).

An economy will gradually slow capital accumulation as it approaches its steady state. In the steady state, Δln(ItYt)=0. Denoting Δln(At)= g and Δln(Lt)= n, the expression results in: Δln(Kt)=β1αg+n. Using the capital accumulation equation and substituting for Δln(Kt)=ItKt1δ and β = (1 − α), we obtain the steady-state golden rule investment rate in the Solow-Swan growth model with labor-augmenting technological progress:

IY=(g+δ+n)KY

This suggests that our approximation of the transition path is a plausible transition dynamic, since it converges into the balanced growth path.

In order to evaluate the parameters c, α, and β, we use the historical experiences of countries in Western Europe with their capital accumulation path over 1951–2011. Fitting the above specified transition path for the investment rate on a panel for Germany, France, Italy, and Spain over 1951–2011 (R2=0.87, asterisks denote statistical significance with *** at 1 percent and ** at 5 percent), yields:

ln(ItYt)=0.18**+0.7***ln(At)0.6***ln(KtLt)

Using these parameters and a CESEE country-specific K/L ratio and labor-augmenting productivity, we can compute sustainable “historical benchmark” investment rate which mimics earlier transition dynamics of advanced economies.

Annex VII. The Effect of Structural Factors on Total Factor Productivity14

Methodology

Stochastic frontier models are used to analyze the efficiency of economic agents, regions, or countries. The intuition behind the models is that frontier technology may not be exceeded by any of the economic agents and the distance from the frontier reflects the inefficiency of individual agents. The frontier represents the maximum amount of output that can be obtained from a given level of inputs. Stochastic frontier models are characterized by composite error that is composed of idiosyncratic disturbance (to capture measurement errors and other noise) and one-sided disturbance, which represents inefficiency. In this annex we use a stochastic frontier panel-data model proposed by Battese and Coelli (1995) to estimate the contributions of technological progress and country-specific technical efficiency to total factor productivity (TFP) growth.15 Stochastic frontier models could be described by the following equations:

yi,t=a+xi,tβ+εi,t(1)
εi,t=vi,tui,t(2)
vi,tN(0,σv2)(3)
ui,t=zi,tδ+wi,t(4)

where yi,t-is the output of country i at time t; xi,t is a vector of production function inputs (in our case, capital, K, and human capital augmented labor, LHC, and the time trend representing technological change; εi,t is the composed error term; vi,t is assumed to be iid random error, independently distributed from the ui,t; ui,t is non-negative random variables associated with the technical inefficiency of production, which are assumed to be independently distributed, such that they are obtained by truncation (at zero) of the normal distribution with the mean; zi,tδ, and variance, σu2,zi,t are a vector of explanatory variables associated with the technical efficiency of production of country i, at time t; δ is an (m x 1) vector of unknown coefficients; and wi,t is defined by the truncation of the normal distribution with zero mean and variance, σu2, such that the point of truncation is zi,tδ i.e., wi,tzi,tδ. These assumptions ensure non-negativity of ui,t. Parameters of the stochastic frontier and the model for the technical inefficiency effects are simultaneously estimated with a maximum likelihood method.

Kumbakhar and Lovell (2000) demonstrate that a change in the TFP, which is defined as output growth not explained by input growth, can be expressed as:

ΔTFP=ΔTP+ΔTE+(1)[lhcΔLHC+kΔK],(5)

where, ΔTP is technological change, which is represented by the coefficient of the time trend in equation (1) of the production frontier; ΔTP is the change in technical efficiency; lhc and k are output elasticities with respect to human-capital-augmented labor and capital, respectively; and = lhc + k represents the return to scale. In the case of constant return to scale, = 1 factor accumulation does not have any impact on TFP growth.

We use stochastic frontier analysis to estimate the production frontier and technical inefficiency, and to identify structural, regulatory, and institutional factors that are associated with technical inefficiency. The analysis applies the stochastic frontier method to 30 advanced and emerging European economies and the United States for the period 1995–2014. 16 The model is estimated using purchasing power parity-adjusted annual data from the Penn World Tables (PWT) and structural variables from the World Bank’s Global Competitiveness Report and from the Economic Freedom of the World Survey. 17 The production function approach is used to remove cyclical components from output and labor series (for a detailed description, see Podpiera, Raei, and Stepanyan, forthcoming). Structural variables cover the following broad areas: (1) product and labor markets, (2) institutional quality, (3) quality of infrastructure, (4) innovation and R&D, and (5) quality of labor and capital.

Regression Results

According to our estimation, technology progressed at 0.5 percent per year, on average, during 1995–2014 (Annex Table VII.1). However, before the global financial crisis the average growth rate of technological progress was higher, at about 1 percent, while after the crisis technological progress stalled. Estimated coefficients for physical capital and human-capital-augmented labor in the production function are very close to the calibrated labor and capital shares used in the literature. These results are robust to the different model specifications and different samples.

Annex Table VII.1.

Estimates of Production Function and Efficiency Components

article image
Standard errors are in parentheses. *** (**) (*) Denotes significance at 1% (5%) (10%) level.Source: IMF staff estimates.Note: FDI=Foreign direct investment.

Differences in structural factors explain the variation in inefficiency across countries in our sample. The high levels of corruption and restricted business regulations, including for foreign direct investment (FDI), give rise to technical inefficiencies. The higher share of employment in the services sector and the longevity of the population are conducive to technical efficiency. The square of human capital has a positive sign on inefficiency, reflecting the diminishing return on human capital.18

We zoomed in to identify specific factors that influence technical inefficiency and, according to the literature, are behind corruption. We used a variety of indicators representing the legal and judicial system in lieu of corruption indicators in our analysis.19 Data limitations prevent us from including all variables simultaneously, and therefore, we used one at a time. The results suggest that judicial independence, impartiality of the courts, and property rights play an important role in improving technical efficiency.

Implications for Structural Reform Priorities

Structural reform priorities vary across countries depending on potential relative efficiency gains (Annex Figure VII.1):

Legal system and protection of property rights: Among CESEE countries, Bulgaria, Croatia, the Slovak Republic, and Slovenia have relatively large room to increase efficiency by improving the legal system (independence of the judicial system and impartiality of courts) and protection of property rights. Turkey also has significant room to improve the independence of the judicial system. Albania, Hungary, Serbia, and the CIS countries could also benefit from improving protection of property rights. In general, the Baltic countries have institutional and structural characteristics very close to the EU-15 average. Thus, the room to gain efficiency by improving these characteristics is limited.

Annex Figure VII.1.
Annex Figure VII.1.

Potential Efficiency Gains from Improving Selected Structural Characteristics of CESEE Economies to the Average EU-15 Level

(Percent)

Sources: World Bank; Global Competitiveness Report; Economic Freedom of World; and IMF staff calculations.Note: CEE = Central and Eastern Europe; CESEE = Central, Eastern, and Southeastern Europe; CIS = Commonwealth of Independent States; SEE = Southeastern Europe; SEE-XEU = Southeastern European countries outside the EU.

Business regulation: Croatia, the Czech Republic, and the Slovak Republic could gain the most among CESEE countries from easing general business regulation and restrictions for FDI.

Structural transformation: In Albania, Romania, Turkey, and, to a lesser extent, Poland, there is scope to raise productivity by shifting labor from relatively lower productivity sectors (agriculture) to higher-productivity (services) sector.

Life expectancy: the Baltic and the CIS countries have the greatest room to improve the life expectancy of the population.

Annex VIII. Is There a Role for Structural Policies in Improving Allocative Efficiency?20

Methodology

This annex analyzes the role of structural policies in improving the efficiency of resource allocation. The evidence from Organization for Economic Cooperation and Development (OECD) countries, as shown in Andrews and Cingano (2014), suggests that policy-induced frictions in labor, product, and credit markets have an economically and statistically significant negative relationship with aggregate productivity, as they can hinder efficient resource allocation from less to more productive firms. This annex applies a similar methodology to 14 CESEE countries using firm-level data from ORBIS for the period 2010–13, and examines whether certain reforms could help close productivity gaps by facilitating more efficient resource allocation.21

In order to test the role of the quality of institutions and regulations in resource allocation, we estimate the following fixed-effect model with time dummies:

AEi,c,t=mβm×Rc,tm+μi×c+μt+εi,c,t,

where AEi,c,t denotes allocative efficiency, measured by the covariance between firms’ labor productivity and their labor share within industry i of country c, and year t;22, 23 Rc,tm denotes the country-level m-th indicator of regulation and institutional quality; μixc denotes the fixed effects for industry and country groups; and μt denotes time dummies. For structural indicators, we use the World Economic Forum’s Global Competitiveness Index, particularly in the areas of government efficiency, flexibility in wage determination, and affordability of financial services.24

Results suggest that the quality of institutions matters for allocative efficiency, and, in CESEE, the improvement in government efficiency and affordability of financial services could yield significant potential productivity gains through better resource allocation. The regression results suggest that, for instance, an increase in the affordability of financial services indicator by one notch is associated with a rise in allocative efficiency by 13 percentage points. The productivity gains from reforms in these areas (government efficiency and affordability of financial services) combined to bring them up to the higher level observed in the benchmark case of Sweden can be sizable—between 10 and 20 percent depending on a country’s gap from the benchmark (Annex Table VIII.1)

Annex Table VIII.1.

Allocative Efficiency and Structural Indicators

article image
Sources: ORBIS; and IMF staff calculations.Note: Allocative efficiency is measured as the covariance between a firm’s labor share within industry and its log productivity. Robust standard errors are in parentheses. The coefficients on fixed effects and year dummies are omitted. *** coefficient significant at 1 percent; ** significant at 5 percent; * significant at 10 percent.

Data Description

The allocative efficiency analysis uses firm-level data from the ORBIS database, covering over 1.5 million firms each year for the period between 2010 and 2013 (1.8 million firms for 2013) for 14 countries. The sample excludes the self-employed (firms with one employee) and the outlier firms at the top and bottom 1 percentile in terms of their productivity. Allocative efficiency is calculated using each firm’s labor productivity and labor share within industry (using a narrow classification according to NACE Rev. 2, first two-digit level). Those industries with less than 20 firms available are excluded from the sample. Annex Table VIII.2 shows the data coverage (comparing the total number of employees hired by sample firms to the aggregate-level employment data, excluding the finance and insurance sector) and the number of firms and industries for each country.

Annex Table VIII.2.

Number of Observations and Data Coverage by Country, 2013

article image
Sources: Eurostat; and ORBIS.Note: Coverage is the share of total employment by sample firms to aggregate-level nonfinancial sector employment. The sample includes firms with more than one employee.

The data contain a large number of micro firms (with less than 20 employees) that accounts for only a small fraction in terms of employment and turnover—we also examined allocative efficiency based on a subsample excluding such micro firms (Annex Figure VIII.1). The table shows the size of the subsample as a share of the full sample in terms of number of observations, employment, turnover, and total assets.

Annex Figure VIII.1.
Annex Figure VIII.1.

Average Firm Productivity by Firm Size

(Log of turnover-to-employment ratio, weighted by employment)Sources: ORBIS; and IMF staff calculations.CEE = Central and Eastern Europe; CIS = Commonwealth of Independent States; SEE = Southeastern Europe; SEE-XEU = Southeastern European countries outside the EU.

Annex IX. Decomposing TFP Growth into Common and Idiosyncratic Components25

Productivity growth has slowed across countries regardless of their level of development (Eichengreen, Park, and Shin 2015). A widespread slowdown in total factor productivity (TFP) growth raises a natural question: Are common factors behind this slowdown? This annex describes the framework that we used to decompose TFP growth into common/external and country-specific/idiosyncratic factors. For this purpose we run the following regression for each CESEE country separately:

ΔTEPt=α0+α1ΔTFP_COMt+α2ΔTFP_PARt+εtΔTEP_COMt=1Ni=1NΔTFPi,t,i=1, N

where ΔTFPt is TFP growth at time t; ΔTFP_PARt is weighted average TFP growth of trading partners at time t (weighted by exports); ΔTFP_COMt is average TFP growth across countries in the sample at each point in time, which represents other common factors for TFP growth; N is the number of countries in our sample; εt is the country-specific component of TFP growth; and as are parameters that need to be estimated. Vectors of a1 and a2 across all countries in our sample represent common factor loading vectors. To control for country fixed and time effects, all data are de-meaned and de-trended in advance. TFP growth data for CESEE countries are from Podpiera, Raei, and Stepanyan (forthcoming). TFP growth for trading partners is calculated using the production function approach described in Podpiera, Raei, and Stepanyan (forthcoming).

Annex X. Description of Variables

article image
article image
article image
article image