Republic of Moldova: Selected Issues

The first section of this paper is an attempt to examine the interest rate channel of monetary policy transmission in Moldova and to estimate the strength and the speed of the interest rate pass-through. The next section provides a background on Moldovan financial markets, liquidity conditions, and the current framework of monetary policy. The following section sets out the formal model used to estimate the strength and the speed of the pass-through, and the last session discusses results.

Abstract

The first section of this paper is an attempt to examine the interest rate channel of monetary policy transmission in Moldova and to estimate the strength and the speed of the interest rate pass-through. The next section provides a background on Moldovan financial markets, liquidity conditions, and the current framework of monetary policy. The following section sets out the formal model used to estimate the strength and the speed of the pass-through, and the last session discusses results.

III. Changes and Flexibility in the Moldovan Labor Market39

A. Introduction

79. Moldova is in the process of economic and labor market restructuring, increasingly mirroring earlier transition economies. The process has entailed a net loss of about a quarter of the workforce and has had a profound impact on the labor market. While increasing the supply of labor in some sectors of the economy and thus containing wage growth, it has also led to substantial emigration and higher productivity.

80. As the reallocation of labor among sectors is increasingly coming to an end, real wage growth can be expected to increase more rapidly in the future—speeding up both the convergence process as well as the urgency to ensure flexibility in the labor market. Our analysis shows that while overall real wage developments are broadly in line with fundamentals, emigration has functioned as a short-term “shock-absorber” for the economy as a whole and for the industrial sector in particular. As the labor market gradually tightens in the future, the short-term wage setting mechanism will probably change making it essential to preserve flexibility of wages in order to ensure a close link with productivity.

81. Population and labor statistics have been substantially revised following the 2004 population census, improving data reliability but complicating an in-depth analysis of recent trends. While the methodology, training and practices of surveyors have improved, the most substantial revision was the expansion of sampling, thus achieving a more accurate reflection of population dynamics as well as the labor market. As no comprehensive attempt has been made to revise the old series based on new data, there is a break in the sample for some of the variables used in this analysis.

B. Some Stylized Facts

82. The ongoing restructuring of the Moldovan economy is similar to what other transition economies experienced some years ago (Figure III.1). Throughout the 1990s, agriculture remained GDP above 25 percent of GDP, while other sectors saw only modest changes. Since then agriculture’s share of GDP has fallen by half, while services and construction have expanded rapidly. Industry has remained stagnant as a share of GDP, but this masks substantial intra-sectoral changes as traditional industries have declined while especially export-driven manufacturing industries have expanded.

Figure III.1.
Figure III.1.

Economic Structure

(Shares of GDP, Value Added)

Citation: IMF Staff Country Reports 2008, 134; 10.5089/9781451825145.002.A003

Sources: National Bureau of Statistics, EBRD and Eurostat.* SEE-4 = Serbia, Albania, Bosnia, and Macedonia FYR.

83. Moldova’s population dynamics and participation in economic activity have changed as well (Figure III.2). In the 1990s, the natural rate of growth of population was positive, emigration levels were relatively low, and economic participation levels were high. More recently birth rates have sharply declined and emigration has increased leading to an 8 percent fall in population since 2000. Partly because most emigrants are mainly of working-age, the economic participation rate (labor force as share of working age population) fell sharply and has now reached Romania’s level. Eventually the rate might improve slowly—as seen in some of the early transition countries—as the lower birth cohorts enter working age.

Figure III.2.
Figure III.2.

Population and Overall Labor Market Developments

Citation: IMF Staff Country Reports 2008, 134; 10.5089/9781451825145.002.A003

Sources: National Bureau of Statistics and Eurostat.

84. Developments in labor market closely mirror the transformation of the real economy (Figure III.3). The allocation of labor resources was at the outset somewhat different from other countries in the region with a larger share of agricultural employment. Starting in 2002, the importance of agriculture fell sharply and reached Romania’s level in 2007 (which is similar to Bulgaria’s 2004 level). In contrast to other transition economies the transformation of the labor market, in particular some 350,000 jobs lost mostly in agriculture (one of quarter of the workforce in the year 2000) did not lead to persistently high unemployment levels. This is probably because job opportunities abroad—helped by linguistic and cultural links with larger, close neighboring economies—absorbed the temporary surplus of labor resources. As labor allocation across economic sectors is increasingly becoming similar to other countries in the region, the labor market may gradually begin to tighten. This may in turn put pressure on wages and further slow down emigration.

Figure III.3.
Figure III.3.

Labor Market Developments

Citation: IMF Staff Country Reports 2008, 134; 10.5089/9781451825145.002.A003

Sources: Eurostat and National Bureau of Statistics.

85. The labor market dynamics show a similar trend (Table III.1). While the agriculture sector has seen a stagnant job turnover with sharply falling employment levels, the construction and service sectors have seen the opposite trends. The job turnover in the industrial sector has sharply increased, probably reflecting the replacement of traditional industries (such as wine production, sugar, and tobacco) with newer industries.

Table III.1.

Table Labor Market Dynamics

article image
Source: Labor Force Survey.

The 2006 employment growth was impacted by the Russian embargo on wine import from Moldova.

Defined as the gross number of hired and dismissed employees during the year as a proportion of total employment at the beginning of the year.

86. There have been substantial changes in several segments of the Moldovan population and labor market (Table III.2). The biggest change is the decline in the labor force; whereas about a quarter of the job losses (in net terms) and reduced unemployed have been absorbed by other sectors of the economy, most of the remaining workers have sought employment outside the country. There has also been a substantial increase in the number of economically inactive (defined as working work less than 20 hours a week). For the youngest segment of the labor force (between 15-24 years), this is probably mostly linked to larger higher education enrollment, as both increased by about 69 thousand. For the segment above 65 years, about 41 thousand are no longer counted as employed, leaving some 35 thousand in this segment in the labor force. Despite the more than 187,000 new pensions granted during 2000-06 (Table III.3), the number of economically inactive in the 50-64 age group increased by only 40 thousand. This indicates an increase in economically inactive of more than 100 thousand people between 24 and 50 years; these might be supported by seasonal migration or through remittances indicating a higher reservation wage.

Table III.2.

Moldova: Changes in Population and Employment, 2000-06

(In thousands)

article image
Source: National Bureau of Statistics.

87. Labor productivity has substantially increased (Figure III.4). Not surprisingly, the largest (per worker) productivity gains have been in agriculture as output has increased and the employment level has fallen sharply. Similar to many other countries in the region, the real wages increases have in general been larger than the increases in productivity. However, except for the industry and construction, the wage increases in Moldova have been remarkably similar in all sectors, perhaps a result of the wage setting mechanism (see below).

Figure III.4.
Figure III.4.

Labor Productivity and Real Salaries, 2000-06

Citation: IMF Staff Country Reports 2008, 134; 10.5089/9781451825145.002.A003

Source: World Economic Outlook and National Bureau of Statistics

88. The increased gap between wages in Moldova and those in neighboring countries despite the higher productivity gains is indeed a puzzle (Figure III.5 and Annex 1). The reason might be that the “excess” of workers in traditional sectors such as industry and agriculture have moved into other sectors of the economy, thereby depressing wage growth. This however, might be shifting as there is anecdotal evidence of labor shortages in certain sectors and sporadic attempts to attract Moldovans working aboard. However, while some sectoral labor shortage may occur as new industries develop (such as the ongoing expansion in call-centers), any tightening of the labor market will most likely be gradual; the recent increase in working age economically inactive as well as that agriculture still accounts for a substantial percentage of those employed indicate a sufficient short- to medium-term supply of labor. However, whether these represent a readily available pool of workers for more modern sectors of the economy is still unclear.

Figure III.5.
Figure III.5.

Difference between Monthly Moldovan and Average Regional Wages, by Sector

(In US dollars; Region = Bulgaria, Romania, Russia, and

Citation: IMF Staff Country Reports 2008, 134; 10.5089/9781451825145.002.A003

Sources: National statistical agencies.

89. The recent emigration of workers does not appear to have substantially impacted growth prospects (Figure III.6). Emigrants are primarily men with secondary education from rural areas, representing around 20 percent of that segment of the 2004 population. The depletion of human capital appears to be relatively modest partly because university educated people overwhelmingly seek domestic employment. As only about one third of the emigrants tend to remain abroad, a large portion of the population has international work experience, some of which could potentially increase productivity of the domestic workforce.42 The outlook is less clear, as the increasing gap between salaries in Moldova and neighboring countries can increase emigration and seriously undermine future growth. On the other hand, the wage gap has increased competitiveness and probably explains part of the recent surge in investments in export-related industries and in foreign direct investments.

Figure III.6.
Figure III.6.

Moldova: Share of 2004 Population Abroad in 2006, by Demographic Category

Citation: IMF Staff Country Reports 2008, 134; 10.5089/9781451825145.002.A003

Source: Labor Force Survey.

C. Institutional Framework of the Labor Market

90. Many institutional factors of the labor market impact worker incentives. In transition economies in particular, these supply-side factors can impact the level of informal employment, the size of the shadow economy, migration, as well as the competitiveness of the overall economy. The factors that appear to be most relevant to Moldova are discussed below.

The wage setting process

91. The private sector wage setting process43 in Moldova appears to be largely decoupled from economy, sector- or firm-level productivity. The process takes place in a tripartite framework in which the government plays a decisive role through a traditional system of general wage indexation based on the statutory minimum wage.44 Although this so-called “tariff schedule system” has been formally abolished, it reportedly remains more common than bilateral firm- or sector-negotiated agreements, at least in traditional industries and sectors. The collective agreements between the social partners on the national, regional, branch and enterprise levels that do take place typically focus on separate wage adjustments to the indexing system. This may explain why the wage increases in Moldova have been remarkably similar in all sectors, despite the varying productivity gains. The predominance of the tariff schedule system, which appears to be similar to the “wage tariff system” in the Slovak republic (IMF, 2006) might be reduced in the future through legislative45 and other changes (such as that newer industries and firms seem to be increasingly using a more decentralized wage setting mechanism).

The tax wedge

92. The tax wedge46 in Moldova was 32 percent in 2006, relatively low by regional standards (Figure III.7). This could indicate that the potential financial disincentive effects on labor effort are relatively modest in Moldova. Moreover, the tax wedge is estimated to have fallen further to below 31 percent in 2007, and will probably remain unchanged in 2008 as the lowered tax rates are offset by modestly higher contributions (IMF, 2007).

Figure III.7.
Figure III.7.

Tax Wedge in Selected Countries47

(Percent of Gross Wages, 2006)

Citation: IMF Staff Country Reports 2008, 134; 10.5089/9781451825145.002.A003

Source: World Bank (2007).

93. Personal income taxes play a less important role than contributions in determining the wedge. Three-fourths of the wedge is explained by social security contributions. The second biggest factor is the health fund contributions, which explain 13 percent of the wedge. Personal income tax is the third factor and explains the remaining 12 percent of the wedge (IMF, 2007). However, the role of contributions might be overstated as enforcement is generally considered to be weak. The reason might be that according to current legislation, employees are entitled to benefits irrespective of actual employer contribution levels.

94. Moldovan personal income tax rates48 are also relatively low compared to other countries in the region. The best comparison is probably the effective tax rate (personal income tax over the average yearly wages) which is quite low in Moldova (Figure III.8). Comparing the top marginal rates might not be very useful since countries differ (for example, different level of deductions and income distribution of taxpayers), but indicates that the Moldovan 20 percent rate (reduced to 18 percent in 2008) is relatively low (e.g., Poland has 40 percent, Hungary 36 percent, Slovenia 41 percent), although it is obviously higher than in flat-rate countries like Russia (13 percent), Romania (16 percent), Macedonia and Serbia (12 percent).

Figure III.8.
Figure III.8.

Personal Income Tax Rate (effective), 2006

Citation: IMF Staff Country Reports 2008, 134; 10.5089/9781451825145.002.A003

Source: World Bank (2007).

Social benefits

95. Social benefits do not appear to be a deterrent to employment in Moldova. The main reason for the low level of unemployment benefit recipients might be the social security contribution requirement, disqualifying about one third of the labor force that are informally employed (ILO, 2005). Even for those that do qualify, the level of unemployment benefit is quite low, despite the legal obligation that the minimum level of the unemployment benefits should be equivalent to 50 percent of the national average wage. In addition, there is not yet any functioning targeted social assistance system in place and active labor policies seem to play a very minor role in Moldova (Table III.3). As a consequence, any notion of “unemployment trap” or “low income trap” do not seem to be particularly relevant (IMF, 2006); instead the low level of social benefits could help explain the relatively low level of unemployment and part of the emigration in Moldova.

Table III.3.

Social Benefits, 1999-2006

article image
Sources: Ministry of Economy and Trade of the Republic of Moldova and National Agency for Employment.

96. The recent sharp increase in new pensions does not appear to have substantially reduced the incentives to work. Unlike in other transition countries, the use of pensions to mitigate the effects of economic restructuring was not as widespread during the 1990s in Moldova as in other transition economies (Fortuny, Nesporova and Popova. 2003). As noted above, despite that the number of new pensions granted every year have almost tripled since 1999, it has had very little impact on the number of economically inactive among the population receiving the pensions. The reason might be that pension benefits are still so low that unless coupled with other sources of income, pensioners in general fall below the poverty line.49 This implies that a substantial increase in pensions might abruptly tighten the labor market.

Employment protection legislation

97. Moldova’s employment protection legislation remains significantly more rigid than the OECD average. Although a new Labor code was adopted in 2003, certain provisions such as redundancy obligations (substantially larger severance pay obligations for employees with long seniority50 and third party notification) drag down the score on rankings such as the World Bank’s “Doing Business” survey (Table III.4). However, the high rate of job destruction suggest the that these obligations are not binding constraints for the labor market (Ristowski, 2004).

Table III.4.

Employing Workers Indicators, 2008

article image
Source: World Bank Doing Business project, 2008.

The Rigidity of Employment Index is the average of three sub-indices: Difficulty of Hiring Index, Rigidity of Hours Index, and Difficulty of Firing Index. Each of these indices assigns values between 0 and 100, with higher values representing more rigid regulations.

D. Flexibility in the Labor Market—Empirical Analysis of Determinants of Real Wages

98. This paper uses an econometric analysis to assess the degree to which the labor market has been able to adjust to recent changes in the economic situation. To formally test the degree of flexibility of real wages, we use the Zovas and Melihovs (2005) approach, which specifies that real wages are determined by changes in productivity and unemployment. However, to fully assess the level to which Moldova’s wages have performed as a shock absorbing mechanism, we have included minimum wages (given the indexation of wages) and emigration as only about one third of the recent shock has been absorbed by the domestic economy, the rest (in net terms) have sought employment outside the country. The factors included in the analysis are briefly described below:

  • Minimum wages. The impact of minimum salaries on real wage wages—and indeed unemployment—depends on whether it is set at market clearing levels. In Moldova, minimum salaries are currently set at 900 lei/month for the general economy (at about one third of average monthly salaries)51 and it is doubtful that many receive this salary only. However, as discussed above the main impact of minimum salaries on real wages may be indirect through the “tariff-schedule system”. If this is applied widely, minimum wages are expected to have a positive impact on real salaries, possibly distorting the relationship between productivity and real salaries.

  • Productivity. According to the marginal productivity theory of wages, the highest wage an employer is willing to pay an employee is equal to the additional value that one extra worker generates. Consequently, with labor productivity increasing in the economy, a real wage increase can be anticipated. This theory deals wage setting mechanism only from the demand side in that wages are determined by the amount an employer is willing to pay. In Moldova, the wage setting mechanism might distort the relationship between real wages and productivity.

  • Emigration of workers reduces the labor supply in the economy or at least in the sectors that are affected. Hence, workers that stay behind stand to gain in terms of real wage increases unless there are offsetting factors in the economy. As noted above, Moldova has experienced substantial emigration, especially as agriculture employment has declined, and to the extent this has reduced labor supply it can be expected to be positively correlated with real salaries in the economy.

  • Unemployment can be viewed as an indicator of equilibrium in the labor market. When unemployment falls, it might indicate an increase in demand for labor and hence employers view marginal product of labor to have increased which in turn can have an impact on real wages. Given that the unemployment level in Moldova has remained below the regional average throughout the recent changes in the economy and the labor market, it is doubtful that the unemployment rate has been a decisive factor in the wage setting process in Moldova.

Econometric results

99. The long-run and short-term results below are from a vector error correction model, using a monthly dataset from 2000-2006. All variables used in this analysis were tested for seasonality, and those that were found to be seasonal were adjusted using the X-12 ARIMA methodology. Using the augmented Dickey-Fuller (ADF) test, all variables were found to be stationary in first differences, and all equations had at least one cointegrating vector. Most variables were unaffected by the change in methodology (and those that were, have been adjusted using the trend in the old data series), hence a dummy variable was deemed, and indeed found to be, unnecessary.

Determinants of long run real wages

100. The following equations were obtained describing the long-term wage setting mechanism in the overall economy and some specific sectors:

article image

where all variables are in logs and RW is real monthly wages, RMINWG is the sectoral minimum wage deflated by CPI, RPROD is real increases in value added, EMIGR is the total stock of emigrants, and UEMP is the deviation of unemployment from its long-run trend.

101. The results seem to broadly confirm that the specified factors help determine as expected long run real wages in the economy as whole. The estimated coefficients show that a 1 percent increase in real minimum wages tend to increase real wages by 0.5 percent, a 1 percent increase in productivity leads to a 0.9 percent increase, and finally that a 1 percent increase (decrease) in unemployment from it’s long-run trend decreases (increases) real wages by 0.1 percent. A possible interpretation of these results might be that despite widespread indexation of wages, there has overall been a close correlation between productivity and real wages. While unemployment has only played a very minor role in determining long run real wages during the labor reallocation period, emigration has had no impact as it has mainly (in net terms) been a channel for “surplus” labor from the traditional sectors.

102. The sector-specific results show a somewhat different story. The effect of minimum wages for most sectors assessed (which represent just a part of the economy) are similar to those for the economy as a whole, except for construction for which it is negative and significant. The reason for this might be that wage indexation in this sector is not widespread; the impact of minimum wages in other parts of the economy induces layoffs and thus an increase in labor supply for the construction sector (thus putting pressure on real wages). In terms of increases in productivity, only the construction sector appears to be passing on gains in the form of real wages increases; the link in the trade sector appears to be negative perhaps indicating that the large recent increases in output have been caused by more workers (which are mainly informally employed in this sector) at the cost of lower real salaries. Emigration for all sectors assessed show as expected a positive correlation, as the reduced labor supply have put pressures on real wages. Unemployment is only significant for trade and construction; for the latter, higher unemployment seem to be positively correlated with real salary increases perhaps indicating that people tend to register as unemployed instead of migrating when the real wages in the construction sector—which has indeed been the sector creating most new jobs—go up.

Short-term labor flexibility

103. For labor market flexibility analysis, short-term wage fluctuations and adjustments to the long-term trend may be more important than the long-term developments. The results are as follows (only statistically significant results are reported):

article image

104. The results indicate that emigration has been the short term “shock absorber” for the economy during the labor reallocation. The estimated coefficients show that, in the short term for the economy as a whole and in the industrial sector, emigration explains some 9 percent of the variation in real wages. Emigration is not found to be significant in the short term for other sectors, which might indicate that the labor reallocation process has provided sufficient labor supply in these sectors and that supply rigidities such as required skills are not binding in these sectors.

E. Summary and Policy Implications

105. The overall labor reallocation process seems to be coming to an end as the Moldovan economic and labor market structure increasingly mirrors other transition countries. While the process has increased supply of labor in some sectors of the economy and thus helped to contain wage growth, it has also led to substantial emigration and in turn higher productivity. In the future, a gradual tightening of the labor market can be expected, which might put increasing pressure real wages. This increases the necessity to ensure flexibility in the labor market to ensure a close link between real wage and productivity growth. Hence, possible policy implications might be:

  • Reform the wage setting mechanism. In the future, it may no longer be possible to rely on a shift of employment from the agriculture sector to other sectors to contain wage growth. Instead, efforts might be needed to ensure that the wage setting mechanism reflect industry and firm-specific productivity differences. Differentiated wage gains might in turn help close some of the sectoral wages gaps with neighboring countries, and thus further stem emigration.

  • Consider reforms that might impact labor supply. The overall labor supply seems sufficient to facilitate the ongoing labor transformation process, although pension reform might speed up the tightening of the labor market. Efforts should therefore probably focus on mitigating any emerging sectoral labor shortages that might occur, possibly by attracting non-resident workers and ensuring cross-sectoral labor flexibility.

Annex III. 1. Monthly wages, Selected Countries 2000-06

uA03app01fig01
Sources: National statistical agencies.

References

  • European Bank for Reconstruction and Development. 2005. “Transition Report 2005: Business in Transition”, London, UK.

  • Fortuny, M., Nesporova, A. and Popova, N. 2003. “Employment promotion policies for older workers in the EU accession countries, the Russian Federation and Ukraine”, Employment Paper, 2003/50. ILO, Geneva.

    • Search Google Scholar
    • Export Citation
  • International Labor Office and the Council of Europe. 2005. “Country Review of the Employment Policy of Moldova”.

  • International Monetary Fund. 2006. “Republic of Latvia: Selected Issues”, IMF Country Report No. 06/354, International Monetary Fund, Washington, DC, USA.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund. 2007. “Moldova: Reforming the Income Tax”, International Monetary Fund, Washington, DC, USA.

  • National Bureau of Statistics of the Republic of Moldova. 2007. “Labor Force in the Republic of Moldova: Employment and Unemployment, 2007”, Labor Force Survey, Chisinau, Moldova.

    • Search Google Scholar
    • Export Citation
  • National Bureau of Statistics of the Republic of Moldova. 2007. “Labor Market in the Republic of Moldova”, Statistical Compilation, Chisinau, Moldova.

    • Search Google Scholar
    • Export Citation
  • National Bureau of Statistics of the Republic of Moldova. 2007. “Statistical yearbook of the Republic of Moldova”, Chisinau, Moldova.

    • Search Google Scholar
    • Export Citation
  • Rutkowski, J. 2004, “Firms, Jobs and Employment in Moldova”, World Bank Policy Research Paper 3253. World Bank.

  • Zasova, A. and Melihovs, A. 2005, “Assessment of Labor Market Elasticity in Latvia”, Working Paper 3/2005, Bank of Latvia.

  • World Bank. 2004. “Recession, Recovery and Poverty in Moldova”. World Bank.

  • World Bank. 2007. “Doing Business 2008: Moldova”, Doing Business Project, World Bank.

  • World Bank. 2007. “Fiscal Policy and Economic Growth. Lessons for Eastern Europe and Central Asia.” Editors: C. Gray, T. Lane, and A. Varoudakis.

    • Search Google Scholar
    • Export Citation
39

Prepared by Johan Mathisen (EUR).

40

De jure population is a concept under which individuals are recorded (or are attributed) to a geographical area on the basis of the place of residence (source: OECD).

41

De facto population is a concept under which individuals are recorded (or are attributed) to the geographical area where they were present at a specified time (source: OECD).

42

Surveys indicate that less than 40 percent of the emigrant population has been abroad for more than one year.

43

In the public sector, wages are directly set by government decree after tripartite consultations process.

44

Although it is unlikely that many workers actually receive the minimum wage (about 80 dollars—about a half of the official subsistence minimum) only, it serves as a reference point for indexing salaries tiers based on type and difficulty of work.

45

While the legal framework for a collective bargaining system has been put in place, it is not yet operational.

46

Tax wedge is here defined as the sum of personal income tax and all social contributions over the total labor cost.

47

Data refer to effective rates on average wage.

48

Until 2008, the personal income tax had a progressive rate structure with 3 brackets (7, 10 and 20 percent). In 2008, the structure was changed to a two-tier system of 7 and 18 percent.

49

The World Bank estimated that 87 percent of those received pension income in 2002 fell below the poverty line (World Bank, 2004).

50

The Moldovan average is 28.7 weeks pay for workers with 20 years of employment.

51

For the agriculture sector, the minimum salaries are set at 700 lei per month, also about one third of average salaries in the sector.

Republic of Moldova: Selected Issues
Author: International Monetary Fund