Republic of Moldova: Recent Economic Developments

This paper analyzes the recent economic developments in Moldova by reviewing the real, fiscal, and external sectors developments; money banking; and structural policies. The study provides a statistical analysis on the composition of fiscal adjustments in the country, and describes the recent trends in social spending and social indicators in Moldova and in other transition economies in the 1990s, the methodology for estimating efficiency in public spending on education and health care, and assesses the current account determination in the country.

Abstract

This paper analyzes the recent economic developments in Moldova by reviewing the real, fiscal, and external sectors developments; money banking; and structural policies. The study provides a statistical analysis on the composition of fiscal adjustments in the country, and describes the recent trends in social spending and social indicators in Moldova and in other transition economies in the 1990s, the methodology for estimating efficiency in public spending on education and health care, and assesses the current account determination in the country.

VI. Current Account Determination in Moldova55

A. Introduction

130. Since independence in 1989, Moldova’s current account balance has fluctuated widely (Figure 19). The sharp deterioration in the country’s external position in 1992-1993 was due primarily to a terms-of-trade shock, as prices on energy imports from Russia and Ukraine soared. The current account worsened again between 1995 and 1997 reflecting a rapid growth of imports, mainly consumer goods; widening fiscal deficits; and an accumulation of foreign arrears. Export performance was weak throughout the period due to poor harvests, structural imbalances in the agricultural sector, and the inability to expand into new markets. In August 1998 Moldova was hard hit by the financial crisis in Russia, its main trading partner, that led to a severe decrease in both exports and imports. In 1999, the country experienced a dramatic external adjustment as the fall in imports more than offset the fall in exports as a result of a drop in income and a major fiscal contraction.

Figure 19.
Figure 19.

Moldova: Current Account Balance, 1993-2000

(In percent of GDP)

Citation: IMF Staff Country Reports 2001, 022; 10.5089/9781451824957.002.A006

Sources: Moldovan authorities; and Fund staff estimates.

131. Moldova has progressively liberalized its trade and exchange regimes over the past decade. Export licensing had been limited and then completely abolished in mid-1997, import quotas were eliminated by 1996, and the tariff system was simplified with the average tariff substantially reduced. Moldova continues negotiations for accession to the WTO and has free trade agreements with all the CIS countries and Romania. The leu was introduced as Moldova’s national currency in November 1993 and enjoyed remarkable stability until August 1998. In early November 1998, the National Bank of Moldova (NBM) had to abandon the regime of managed float in favor of the fully floating exchange rate. The leu depreciated dramatically following the Russian crisis and ensuing political instability in Russia and at home. Moldova accepted the obligations of Article VIII (sections 2, 3, and 4) of the Fund’s Articles of Agreement in June 1995. Since then, Moldova’s exchange system has remained free of any restrictions on current account transactions, while capital account transactions require licenses and/or registrations from the NBM.

132. Debt instruments have been the most important sources of current account financing. Moldova’s current account deficits have been financed mainly through the accumulation of foreign debt, including borrowing from multilaterals, arrears to energy suppliers, issuance of Eurobonds, sales of treasury bills to nonresidents, and other commercial borrowing. Foreign direct investment has been modest; portfolio investment has been negligible. Large capital outflows (short-term capital flight, as well as debt amortization) took place in the aftermath of the Russian crisis. 1999 was marked by an impressive reversal in Moldova’s current account position.

133. The observed volatility of Moldova’s current account reflects the country’s access to foreign sources of finance. As in other developing and transition economies, excessive volatility of the current account as compared to that of the national cash flow (national income net of private investment and government spending) would indicate imperfect capital mobility. In order to estimate the relationship between current account imbalances and capital flows in Moldova over time, we focus on the recent consumption-smoothing literature on international capital mobility and current account determination (Ghosh, 1995; Glick and Rogoff, 1995; Ostry, 1996).

134. This chapter is organized as follows. Section II briefly reviews the literature on current account determination. Section III describes the methodology used in this chapter. Section IV presents the data. Section V reports the empirical results and Section VI concludes.

B. Brief Literature Review

135. The international capital mobility literature offers several options to test the capital mobility hypothesis. They are:

  • The traditional approach pioneered by Feldstein and Horioka (1980) consists of regressing the investment ratio on the savings ratio for a sample of countries. A positive, close-to-one coefficient of the savings rate is suggestive of imperfect capital mobility. Several economists have challenged Feldstein and Horioka’s interpretation, on the grounds that the high correlation of national savings and investment may be the result of a number of plausible macroeconomic factors (see Golub, 1990; Obstfeld, 1986; Hussein, 1998).

  • Alternatively, the capital mobility hypothesis can be tested in terms of deviations from international parity conditions: covered interest parity (CIP), uncovered interest parity (UIP), and real interest parity (RIP) (Moosa, 1996). The argument is that, if capital is perfectly mobile, then its rate of return should be equal across countries. Satisfactory testing of this argument nevertheless depends upon finding perfectly substitutable assets located in different countries (Mishkin, 1984; Frankel and MacArthur, 1988; Taylor, 1987), which is not easy.

  • A more recent approach, and the one followed here, consists of assessing capital mobility in terms of consumption smoothing in response to shocks to domestic expenditure variables. Ghosh (1995) tests whether capital has been sufficiently mobile in five industrialized countries (United States, Japan, Germany, United Kingdom, and Canada), by comparing the variance of actual and optimal consumption-smoothing current account balances. Deviations from the optimal value indicate imperfect capital mobility. Ghosh finds that capital flows have been more volatile than expected changes in the national cash flow in his sample of industrial countries, except for the United States, and attributes excessive volatility to capital market barriers that prevent optimal international risk-sharing. Evidence for developing and emerging economies is provided by Hussein and de Mello (1999).

C. The Methodology

136. This chapter follows a two-stage approach. First, we estimate the degree of capital mobility in Moldova using a VAR representation of the current account dynamics and test a number of restrictions consistent with the hypothesis of perfect capital mobility. Second, we estimate how the current account balance responds to exogenous shocks in the national cash flow (national income net of private investment and government spending), and its components. This is because capital flows tend to be more volatile and current account imbalances tend to be larger in developing economies than in developed economies, which suggests that there may be less scope for optimal consumption smoothing and international risk-pooling in the former countries.

The VAR Model

137. Testing the capital mobility hypothesis consists of comparing actual current account balances against the benchmark values obtained under frictionless capital mobility and optimal consumption smoothing. Deviations of actual current account balances from benchmark values are indicative of friction in international capital markets: the larger the deviation, the more imperfect capital mobility.

138. In more formal terms, and following recent methodological developments in testing for capital mobility (Campbell and Shiller, 1987), the following VAR can be estimated:

  • (1)Z(t)=ΠZ(t1)+v(t),

    where Z(t) = [Δy(t) Δi(t) Δg(t) CA(t)]′; Π is the transition matrix of the VAR; y is GDP; i, g, and CA are interest, government spending and the current account (all in percentages of GDP); and v(t) is a white-noise term.

    The transition matrix in equation (1) can be decomposed as Π = αβ′, where α=[αy αi, αg αCA] represents the speed of adjustment to equilibrium and β = [βy βi βg βCA]′ is the matrix of long-run coefficients. In this framework, the hypothesis of capital mobility can be tested following two approaches:56

  • The restricted multivariate approach. The hypothesis that capital flows respond to consumption-smoothing behavior requires the changes in the national cash flow (and/or its components) to be endogenous to the current account balance. This can be tested by imposing the following restriction on the VAR coefficients: α = [1 −1 −1 0].57

  • The bivariate approach. Alternatively, following Campbell and Shiller (1987), the VAR can be estimated using Z¯(t)=[ΔAB(t)CA(t)], where ΔAB(t) = Δ(yig)(t), and imposing the following exogeneity restriction on the loading parameters: α = [0 1].58

The Variance Decomposition Analysis

139. Variance decomposition analysis is used to estimate the impact of exogenous shocks to the national cash flow (or its components) on the current account balance, and the dynamic responses of the system to those exogenous shocks. This is important because developing and transition economies tend to be more shock-prone than their more developed counterparts, and current account disequilibria may reflect the responsiveness of the current account balance to exogenous shocks in the national cash flow, or in one or more of its components (national income, private investment, and net government spending).

140. Variance decomposition analysis is conducted using the following reduced-form, common-trend representation of the VAR system in equation (1):

(2)ΔZ(t)=A(L)v(t),

where A(L) is a square matrix of lag polynomials.

141. The technique consists of orthogonalizing the multivariate VAR residuals, by (Cholesky) decomposing them into as many orthogonal time series as endogenous variables in vector Z(t). In this case, the dynamic responses of the current account balances to exogenous shocks to the national cash flow or its components can be simulated. The policymaker can evaluate which components of the national cash flow have a stronger dynamic impact on the current account balance and therefore assess the vulnerability of the current account to external and domestic shocks.

D. The Data

Data Sources

142. The data sources for this study include information provided by Moldovan authorities and IMF staff estimates. National income data for Moldova are not readily available. In this case, we use GDP as a proxy for national income. The GDP data include staff estimates of the shadow economy in Moldova. The quality of private investment data for Moldova is poor. Therefore, a fixed investment time series is used. The data are available from the National Accounts, but include both private and public investment.59 The time series for GDP, investment, and government spending were converted to US dollars at the market exchange rate (Figure 20).

Figure 20.
Figure 20.

Moldova: GDP, Fixed Investment, and Government Spending, 1993-2000

(In percent of GDP)

Citation: IMF Staff Country Reports 2001, 022; 10.5089/9781451824957.002.A006

Sources: Moldovan authorities; and Fund staff estimates.

The Unit Root Properties of the Data

143. The preliminary step in our analysis is to test the degree of integration of the relevant variables. For this purpose, we use the Augmented Dickey-Fuller (ADF) procedure, where the degree of augmentation is determined by the minimum number of lags to produce white-noise error terms. Because of the structural breaks in the series, we also test for unit roots using the Phillips-Perron (PP) methodology. The tests are summarized in Table 26.

Table 26.

Moldova: Unit Root Tests, 1993-2000

article image
Sources: Data provided by the authorities; and IMF staff estimations.***, **, and * denote rejection of the unit root hypothesis at the 1 percent, 5 percent, and 10 percent levels, respectively.

144. In view of strong seasonally of the data, in what follows all the series were seasonally adjusted. The results indicate that the non-stationarity hypothesis is rejected for the level of the current account balance. For the national cash flow and its components (government spending, investment, and GDP) the null hypothesis of a unit root is rejected when the variables are defined in first differences.

E. The Results

The VAR Analysis

145. The exogeneity test shows that the national cash flow components are weakly exogenous with respect to the current account balance. However, the national cash flow itself fails the exogeneity requirement60 and hence the single-equation analysis would be misleading in that case. Therefore, the analysis has to be restricted to the multivariate case.61 The results of the VAR analysis are summarized in Table 27.

Table 27.

Moldova: VAR Analysis, 1993-2000

(current account equation)

article image
Sources: Data provided by the authorities; and IMF staff estimations.(***), (**), and (*) denote statistical significance of the coefficients at the 1 percent, 5 percent, and 10 percent levels, respectively. p-values in parentheses.

146. The components of the national cash flow (income, investment, and public spending) are found to be important determinants of current account variations over time. Fixed investment and government spending have the correct (negative) sign, consistent with partial foreign financing of investment projects and government expenditure. However, the sign on the change in GDP is negative on the first lag and positive on the second lag, the latter consistent with the cash flow argument. The negative sign on the first lag of the GDP variable is suggestive of a one-off deterioration of the current account balance reflecting, for instance, the surge in imports when economic activity accelerates. Subsequently, the current account balance is likely to improve in response to, for example, a reaction of exports, a switch in consumption demand away from imported goods, or an improvement in the budget.

147. Impulse response functions (Figure 21) show adjustment of the current account (CASA) in response to a one-standard deviation exogenous shock to national income (LYSA), government spending (LGSA), and investment (LISA).

Figure 21.
Figure 21.

Moldova: VAR Impulse Response Functions

Citation: IMF Staff Country Reports 2001, 022; 10.5089/9781451824957.002.A006

148. The current account balance is responsive to exogenous shocks to national income. The dotted lines in Figure 21 are confidence intervals. These shocks tend to deteriorate the current account balance in the first quarter after the shock, as discussed above. The current account balance subsequently improves in the second quarter in the aftermath of the shock. Although the responses of the current account balance to exogenous shocks to government spending and investment are correctly signed, their statistical significance is borderline.

Variance Decomposition Analysis

149. The results show that the current account responds strongly to innovations in national income. In line with the impulse responses estimated above, Table 28 reports the results of the variance decomposition analysis of the responses of the current account balance to exogenous shocks. Nearly half of the dynamic response of the current account is accounted for by changes in GDP after 8 quarters in the aftermath of the exogenous shock, followed by innovations in government spending (6.0 percent) and investment (3.7 percent). The forecast error is nevertheless large, suggesting that the results of these simulations should be interpreted with caution.

Table 28.

Moldova: Variance Decomposition Analysis, 1993-2000

article image
Sources: Data provided by the authorities, and IMF staff estimations

F. Conclusions

150. Recent developments in the literature favor the estimation of current account equations in a VAR set-up. The empirical results presented above show that evidence of capital mobility and consumption smoothing in Moldova is mixed. While components of the national cash flow are weakly exogenous with respect to the current account balance, the aggregate cash flow is not. This finding suggests that while the current account is responsive to shocks in national income, investment, and public spending, a significant offsetting adjustment occurs among these variables, so that the impact of changes in the aggregate cash flow on the current account is suppressed.

151. The empirical findings point out that the current account balance in Moldova is strongly responsive to changes in the national income. The current account deteriorates when economic activity picks up but tends to improve subsequently. The dynamic responses of the current account to shocks in national income tend to last two quarters. The current account response to shocks in investment and government spending is much weaker. The current account tends to deteriorate as more foreign financing goes into domestic investment projects and public expenses.

152. The results reported in this chapter may be influenced by a high share of concessional borrowing in Moldova.62 Capital mobility and consumption smoothing could be substantially restricted if Moldova had to rely on non-consessional financing of its current account. Some caution is also recommended in the interpretation of the results due to data inadequacies. These are common weaknesses of empirical analyses for transition economies and the time series available for the relevant variables are typically not long enough for more sophisticated hypothesis testing.

153. The results described above have a number of policy implications. The dynamic pattern of the current account response to changes in national income suggests that economic growth in Moldova is overly vulnerable to external constraints, as income growth at the first instance tends to give rise to debt accumulation. Over the longer term, therefore, the ongoing efforts to create a pro-growth economic environment in Moldova should focus on reducing the country’s vulnerability to external shocks. To this end, structural reform in the export-oriented sectors should be intensified, including privatization and restructuring in agriculture and agro-processing. Hard budget constraints should also be enforced throughout the economy in order to contain the responsiveness of imports and ensure allocation of resources to their most productive uses.

References

  • Campbell, J. and Shiller, R. (1987) Cointegration and tests of present value models. Journal of Political Economy 93, 106288.

  • Feldstein, M. and Horioka, C. (1980) Domestic saving and international capital flows. Economic Journal 90, 31429.

  • Frankel, J.A. and MacArthur, A. (1988) Political vs currency premia in international real interest differentials: A study of forward rates for 24 countries. European Economic Review 32, 10831114.

    • Search Google Scholar
    • Export Citation
  • Ghosh, A. (1995) International capital mobility amongst the major industrialized countries: Too little or too much. Economic Journal 128, 10728.

    • Search Google Scholar
    • Export Citation
  • Glick, R. and Rogoff, K.S. (1995) Global versus country-specific productivity shocks and the current account. Journal of Monetary Economics 35, 15992.

    • Search Google Scholar
    • Export Citation
  • Golub, S. (1990) International capital mobility: Net versus gross stocks and flows, Journal of International Money and Finance 9, 424439.

    • Search Google Scholar
    • Export Citation
  • Hussein, K.A. (1998) International capital mobility in OECD countries: The Feldstein-Horioka ‘puzzle’ revisited, Economics Letters 59, 237242.

    • Search Google Scholar
    • Export Citation
  • Hussein, K.A. and de Mello L.R. Jr. (1999) International capital mobility in developing countries: theory and evidence, Journal of International Money and Finance 18, 367381.

    • Search Google Scholar
    • Export Citation
  • Mishkin, F. (1984) Are real interest rates equal across countries? An empirical investigation of international parity conditions. Journal of Finance 39, 13451358.

    • Search Google Scholar
    • Export Citation
  • Moosa, I. (1996) A note on capital mobility. Southern Economic Journal 63, 248254.

  • Obstfeld, M. (1986) Capital mobility in the world economy: Theory and measurement, Carnegie-Rochester Conference Series on Public Policy 31, 124.

    • Search Google Scholar
    • Export Citation
  • Ostry, J.D. (1996) Current account imbalances in ASEAN countries: Are they a problem?”, International Monetary Fund, mimeo.

  • Taylor, M.P. (1987) Covered interest parity: A high frequency, high-quality data study. Economica 54, 429438.

Table 29.

Moldova: Gross Domestic Product by Expenditure 1994-2000 1/

(In millions of lei; at current prices)

article image
Sources: Moldovan Department for Statistical and Sociological Research; and Fund staff estimates.

Excludes Transnistria.

Table 30.

Moldova: Gross Domestic Product by Sector, 1994-2000 1/

(In millions of lei at current prices)

article image
Sources: Moldovan Department for Statistical and Sociological Research; and Fund staff estimates.

Excludes Transnistria and additional staff adjustment for shadow economy (see table 34).

Table 31.

Moldova: Agricultural Production by Product, 1994-99 1/

(In thousands of metric tons)

article image
Sources: Moldovan Department for Statistical and Sociological Research.

Clean weight.

Excludes Transnistria.

Table 32.

Moldova: Animal Husbandry, 1994-99

article image
Sources: Moldovan Department for Statistical and Sociological Research.

Excluding Transnistria.

Slaughter weight.

Productivity measures exclude production on garden plots.

Table 33.

Moldova: Agricultural and Industrial Production Indices, 1994-99

(Percent change from same period of the previous year; period averages)

article image
Sources: Moldovan Department for Statistical and Sociological Research.

Seasonally adjusted.

Preliminary data.

Table 34.

Moldova: Industrial Production by Industry, 1994-991/

(Percent change from the previous year)

article image
Sources: Moldovan Department for Statistical and Sociological Research.

Excluding Transnistria.

Table 35.

Moldova: Unemployment, Unpaid Leave, and Part-time Employment, 1994-99

(In thousands; excludes Transnistria)

article image
Sources: Moldovan Department for Statistical and Sociological Research.
Table 36.

Moldova: Nominal Wages in Different Sectors, 1994-991/

(In lei per month; period average)

article image
Source: State Department of Statistics.

Excludes Transnistria.

1996 data includes fishery, forestry and hunting.

1996 data includes transport, warehouses, and communication.

Table 37.

Moldova: Inflation, 1997-2000

article image
Sources: Department of Statistics; and Fund staff estimates.
Table 38.

Moldova: General Government Budget, 1994-20001/

(In millions of lei; unless otherwise indicated)

article image
Sources: Data provided by authorities, and Fund staff estimates.

The accounts comprise the republican government, local governments, extrabudgetary funds and the Social Fund.

First two quarters

Includes land tax, real estate tax, natural resources tax, stats tax, and private tax.

Includes profit remittances from the National Bank of Moldova and privatization revenues.

Includes extrabudgetary funds on a net basis, administrative, military, indexation of deposits, environment, and unallocated.

Includes transfers from the State Budget.

Includes treasury securities.

Table 39.

Moldova: General Government Revenues, 1994-20001/

(In millions of lei; unless otherwise indicated)

article image
Sources: Data provided by authorities, and Fund staff estimates.

Comprises the republican government, local governments, extrabudgetary funds and the Social Fund.

First two quarters of 2000.

Includes land tax, real estate tax, natural resources tax, state tax, and private tax.