This note presents estimates of potential growth and the output gap in Latvia. The estimates suggest that the output has marked below potential in the early 2000s but the output gap becomes positive and large after EU accession. With unemployment still well above its natural level, the output gap is estimated to be negative in 2012, but is expected to narrow gradually and be closed in the next 3–4 years. Potential growth is expected to be substantially lower than in 2002–07.

Abstract

This note presents estimates of potential growth and the output gap in Latvia. The estimates suggest that the output has marked below potential in the early 2000s but the output gap becomes positive and large after EU accession. With unemployment still well above its natural level, the output gap is estimated to be negative in 2012, but is expected to narrow gradually and be closed in the next 3–4 years. Potential growth is expected to be substantially lower than in 2002–07.

I. Potential Growth and the Output Gap in Latvia1

This note presents estimates of potential growth and the output gap in Latvia from different methods. The estimates suggest that output was below potential in the early 2000s but the output gap became positive and large after EU accession. The extent of overheating in the pre-crisis boom is particularly uncertain but staff believes output was about 5–10 percent above potential before the crisis. The output gap became negative and large during the crisis, reaching about -13 percent of potential output in 2009. With unemployment still well above its natural level, the output gap is estimated to be negative at slightly below 2½ percent of potential output in 2012, but is expected to narrow gradually and be closed in the next 3–4 years. Potential growth is nevertheless expected to be substantially lower than in 2002–07 unless structural reforms enable to reduce faster and further the structural level of unemployment and remove bottlenecks that would allow attracting more investment in the coming years.

A. Introduction

1. The concepts of potential output and its associated output gap are central to many macroeconomic policy discussions. In the near term, the level of potential output indicates the capacity of the economy to expand without leading to inflationary pressures. Over the medium term, it determines the sustainable pace of noninflationary output growth. Moreover, potential output is a crucial input to assess a variety of macroeconomic outcomes, such as the stance of fiscal policy or the sustainability of the external balance. For example, assessing whether Latvia fulfills the commitments under the ‘Fiscal Compact’ will depend on an estimate of the fiscal position net of cyclical effects and hence on the output gap estimate.

2. However, potential output is an unobservable latent variable and its empirical counterpart needs to be estimated. Many different methodologies have been used to estimate potential output, each of them encompassing a different precise definition of potential output and entailing advantages and disadvantages. No specific approach can be taken to be “the” correct one and potential output estimates are subject to substantial uncertainty. This uncertainty is probably even larger for countries like Latvia, a transition economy still going through substantial structural changes and coming out of a severe crisis that has likely rendered obsolete a significant part of the economy’s productive capacity. This note presents estimates from various methods, argues in favor of estimates from a variant of the production function approach and uses these estimates to discuss briefly the room for policies to enhance potential growth going forward.

3. Analysts have relied on different methods to estimate potential output. Most methods can be broadly classified in two groups: univariate statistical procedures and methods based on economic multivariate models.

  • Univariate statistical methods - The most popular methods within univariate procedures are filters that isolate high from low-frequency components, such as the Hodrick-Prescott filter (HP), the Baxter and King filter (BK) and the Christiano and Fitzgerald filter (CF). The main advantage of filters is their simplicity as they are theory-free methods that only use statistical information from the series itself. The simplicity comes at the cost of some drawbacks though. Filtered trend estimates suffer from significant end-sample bias, that is, they are subject to significant revisions as data for later dates becomes available (even if historical data are not revised). Moreover, these techniques are based only on a statistical decomposition and ignore relevant economic information, which can significantly bias estimates. For example, using an HP filter on a sample characterized by a prolonged period of decreasing inflation—and negative output gap—as a result of tight monetary policy would result in an underestimation of potential output. Also, in the presence of structural changes that render obsolete part of the economy’s production structure, as it is probably the case in the wake of the recent crisis, relying on filtered potential output would most likely overestimate the positive output gap before the structural break.

  • Economic models - The methods based on economic models entail a multivariate approach based on economic theory, typically exploiting key relationships describing goods and labor markets, the degree of capacity utilization of factors and structural aspects of the economy. Some widely used examples include the aggregate production function approach (PF), several variants of models using the Kalman filter and additional information from the Phillips’ curve or Okun’s law, structural vector autorregression models and multivariate filter models (e.g. Benes et al., 2010). This note focuses on a version of the aggregate PF approach and also reports results from a structural vector autoregression model (as in Blanchard and Quah, 1989; hereafter BQ).2

4. The benchmark model used in this note is a variant of the production function approach. It implies assuming an aggregate production for the economy and comparing the actual level of input factors with their ‘potential’ or cyclically-adjusted level. The building block is a Solow growth model that relies on two factors that drive growth in the supply side of the economy: labor and physical capital. The emphasis on a production function approach reflects both staff view that it represents an adequate framework for Latvia (where, for instance, population dynamics and structural unemployment play an important role in potential labor and potential output estimates) and convenience in terms of comparability, since it is the method also used by the Latvian authorities and the European Commission.

  • To compute historical values of potential output, estimates of potential factor inputs are needed. A simple and widely used approach is to estimate potential factors (labor and total factor productivity) by HP filtering the actual factor series. The problem with this approach is that it shifts the problems of filters highlighted for trend GDP to the trend estimates of the inputs. The benchmark model adopted here relies on estimates of the ‘natural’ rate of unemployment—and the unemployment gap—and, to the extent possible, on a variant of the Okun’s law relationship to decompose trend from cycle in other factor components.3 The natural unemployment rate used here is a time-varying nonaccelerating inflation rate of unemployment (NAIRU) estimated using a Kalman filter and a Phillips’ curve relationship between prices and unemployment (see Technical Appendix). The degree of capacity utilization is used to infer the potential level of total factor productivity.

  • To compute potential output in future years, projections for the cyclically adjusted factor inputs are needed.

  • The PF method can provide useful information on the determinants of potential growth. The method relies, however, on an overly simplistic representation of the production technology and on the approach taken to infer the potential level of factor inputs (i.e. the NAIRU estimate and the approach adopted to infer potential productivity).

B. Results

Pre-crisis period

5. Estimates suggest that output was below potential in the early 2000s. Estimates from the PF approach suggest that output was 2½ percent below potential in 2002–03, consistent with high unemployment and a NAIRU estimate implying an unemployment gap of around 1–3 percentage points (Figures 1 and 3). The estimate from the HP filter is about the same, while output gap estimates from the BK and CF filters are negative but much smaller (Figure 2, left chart).

Figure 1.1.
Figure 1.1.

Actual and Natural Unemployment Rate

(in percent)

Citation: IMF Staff Country Reports 2013, 029; 10.5089/9781475578522.002.A001

Sources: WEO; and IMF staff estimates.
Figure 1.2.
Figure 1.2.

Potential Output, Univariate Statistical Methods

Citation: IMF Staff Country Reports 2013, 029; 10.5089/9781475578522.002.A001

Sources: WEO; Haver; and IMF staff calculations
Figure 1.3.
Figure 1.3.

Potential Output, Production Function Model

Citation: IMF Staff Country Reports 2013, 029; 10.5089/9781475578522.002.A001

Sources: WEO; Haver; and IMF staff calculations

6. The output gap became positive and large after EU accession in 2004. Both filters and the PF methods suggest that output increased well above potential from about the time of EU accession (Figures 2 and 3). The estimate from the benchmark PF approach, that incorporates information from the goods and labor markets through the NAIRU estimate, indicate that the widening of the output gap accelerated in 2006, consistent with price and wage inflation developments (Figure 3).

7. Output was probably about 5–10 percent above potential before the crisis, although the extent of overheating at the pre-crisis boom is particularly uncertain:

  • On the one hand, filter estimates suggest output might have been as much as 20 percent above potential before the crisis. The HP filter suggests that the output gap surged to about 20 percent in 2007. Also, when the PF model is fed with HP filtered input series, the pre-crisis output gap estimate exceeds 15 percent (using the usual value for annual data of 100 for the smoothing parameter; Figure 3, bottom right chart). But using the HP filter to remove cyclical fluctuations is subject to significant end-sample bias: the 2007 output gap would have been less than 5 percent if estimated with real time information (i.e. with data up to 2007), while extending the sample up to 2009 would lead to a gap estimate for 2007 almost 3 times bigger (Figure 2, right chart).

    Also, results from the HP filter are very sensitive to the choice of smoothing parameter: when the parameter is set to 6.25, as suggested by Ravn and Uhlig (2002), the pre-crisis estimated output gap is half the size than when the parameter is set at 100. The BK and CF filters also indicate a very steep widening of the output gap before the crisis, though less pronounced. All filter estimates suggest that the absolute size of the pre-crisis (positive) gap was almost twice the size of the (negative) gap during the crisis.

  • On the other hand, the benchmark PF model suggests the output gap peaked at around 5 percent of potential output before the crisis, implying that the absolute magnitude of the pre-crisis output gap was less than ½ the crisis gap. This is consistent with developments in the labor market: while unemployment went below the NAIRU during the boom, the absolute magnitude of the pre-crisis unemployment gap was less than ½ the size of the crisis gap. The participation rate was also high during the boom, but not extreme. One caveat to this estimate is that the potential TFP series used in the benchmark model might not be completely clean from cyclical factors. Using an alternative potential TFP series (subsequently smoothed with an HP filter; see Technical Appendix) leads to a gap estimate of about 9½ percent. Estimates from the BQ model also point to a positive output gap between 5 and 10 percent just before the crisis.

  • While acknowledging the uncertainty of estimates, staff believes output was significantly above potential before the crisis, but probably in the 5–10 percent range rather than in the 15–20 percent range.

Figure 1.4.
Figure 1.4.

Output Gap and Unemployment Gap

Citation: IMF Staff Country Reports 2013, 029; 10.5089/9781475578522.002.A001

Sources: WEO; Haver; and IMF staff calculations.
Figure 1.5.
Figure 1.5.

Output Gap - PF and BQ Methods

(in percent of potential output)

Citation: IMF Staff Country Reports 2013, 029; 10.5089/9781475578522.002.A001

Sources: WEO; Haver; and IMF staff estimates.

8. The investment boom financed by capital inflows was the main driver of potential growth before the crisis. Growth decomposition from the benchmark PF approach shows that in the early 2000s and especially after EU accession, potential output growth was driven mainly by a foreign financed investment boom. Productivity growth also played an important role, although this might partially reflect underestimated rise in factors’ utilization.

Figure 1.6.
Figure 1.6.

Contribution to Growth of Potential Output

(in percent)

Citation: IMF Staff Country Reports 2013, 029; 10.5089/9781475578522.002.A001

Sources: WEO; and IMF staff calculations

Crisis and recovery

9. All methods suggest that output fell well below potential in 2009–10. The benchmark PF model suggests the output gap reached -13 percent in 2009. Estimates from filters range from -8 to -10½ percent. All methods have the output gap narrowing from 2010 or 2011, except for the BQ model that shows a larger output gap in 2012 than in 2011 (although quarterly gap estimates from the BQ model are decreasing towards end-2012).

10. The labor force and productivity were the main factors dragging potential growth during the crisis. The sharp contraction in potential TFP in 2008–10 contributed significantly to the contraction in potential growth during the crisis. This is consistent with efficiency losses in reallocating factors across sectors—although the drop in TFP might be capturing underestimation of changes in capacity utilization. Emigration flows that intensified at the onset of the crisis affected potential employment and growth in 2007–08. The negative impact of labor to potential growth in 2011–12 responds to an increase in structural unemployment most likely due to “hysteresis”—a transformation of cyclical into structural unemployment as skills of the long-term unemployed depreciate.

11. The estimates suggest that the effect of the crisis on potential output was large. While potential output was growing at about 6½ percent per year during 2002–07, estimates suggest it contracted by 7 percent between 2007 and 2011. Potential growth resumed in 2012 but, under current estimates, its 2007 level would not be attained until 2013–14.

Medium-term projections

12. The estimated unemployment and output gap is still significantly negative but is projected to close in the next 3 to 4 years.

  • The benchmark PF model indicates that output in 2012 is still about 2.3 percent below potential (above 3 percent using the alternative potential TFP series) but would diminish gradually and be closed by 2015–16. The reason for the still negative output gap in 2012 is the estimated slack in the labor market: the unemployment rate is still about 2¾ percentage points above the 12.3 percent NAIRU estimate. While the NAIRU estimate for 2012 is large, it is consistent with the unemployment rate having been historically high in Latvia: the average unemployment rate between 1996 and 2011 was above 13 percent.4

  • According to the HP filter the output gap in 2012 is almost 5 percent and would be closed by 2015. The HP filter output gap estimate for 2017 is positive at about 3 percent (although it is only 0.5 percent if the sample is extended from 2017 to 2020, highlighting the end-of-sample bias of this method).

  • On the other hand, the BK and CF filters suggest that the output gap is already closed or slightly positive already in 2012—which is at odds with an unemployment rate above 15 percent and a sizeable unemployment gap.

13. Going forward, staff expects potential growth to be substantially lower than before the crisis. Staff expects a gradual recovery of investment as FDI inflows pick up, but at a much slower pace than in the pre-crisis years when capital flows to the region were extraordinary. A slower pace of capital accumulation would also be associated with a more modest growth in productivity. Labor is not expected to contribute to potential growth in the coming years. The natural unemployment rate is expected to decrease gradually from its current estimate at 12.3 percent to 10¾ percent by 2015 and to slightly below 10 percent by 2017. But the gradual reduction in the natural unemployment rate barely offsets the expected negative trend in working age population. Altogether, staff projects potential growth to increase to about 3¾ percent in 2014–17.

C. Policies to Increase Potential Growth

14. Staff projections suggest that policies increasing potential employment could help attaining higher potential growth in the coming years. Given current projections for population and structural unemployment, the contribution of the labor factor to potential growth would be almost nil in 2014–17. Two sets of policies could increase the contribution from labor to growth. First, policies aimed at reducing the natural unemployment rate by addressing structural bottlenecks that are keeping long-term unemployment at high levels could provide a boost to potential output. In this sense, findings from the forthcoming World Bank study could be used to: i) design ALMPs that aimed at alleviating skill mismatches in the labor market; and ii) address current disincentives from tax and benefit systems that act as an unemployment trap, reducing the attractiveness of moving from unemployment to employment. In addition, in an effort to reduce net emigration, the authorities are developing plans aimed at assisting emigrants who wish to return. Other demographic policy efforts could help in the longer term (Staff Report, Appendix II).

Figure 1.7.
Figure 1.7.

Tax System Does not Encourage Employment

(Unemployment trap, single person without children)

Citation: IMF Staff Country Reports 2013, 029; 10.5089/9781475578522.002.A001

Sources: Eurostat

15. Attaining higher potential growth will depend on fostering productivity and attracting more foreign investment over the medium term. The growth rate of investment observed during the boom years is not likely to return. Still, stronger investment than currently projected and faster productivity gains could be achieved if structural reforms to address bottlenecks were adopted (see next chapter).

Technical Appendix

Assumptions underlying the Production Function Model:

  • Output - A Cobb-Douglas specification is assumed for output:

    GDPt=TFPtKt(1α)Lt(α), where K is the capital factor, computed as explained below; L is the labor factor and TFP is total factor productivity computed as the Solow residual; the labor share in production (α) is assumed to be 2/3.

  • Labor factor – The actual level of the labor factor is total employment in the economy.1 To estimate potential employment, total employment is decomposed in two components: the labor force and un/employment.

    The potential level of un/employment is proxied by a time-varying estimate of the natural rate of unemployment called the nonacceleraing inflation rate of unemployment (NAIRU). The NAIRU is inferred from a Phillips curve-type regression similar to Gordon (1997) and Boone et al. (2001) using quarterly data from 1996Q1 to 2012Q3:

πt=πte+B(L)(utNAIRUt)+d(ut)+C(L)zt+et

where πt and πte denote realized and expected core inflation2; ut denotes the unemployment rate; zt is a set of variables capturing supply side shocks and normalized to zero (in this paper, changes in consumer taxes and variations in import prices); B(L) and C(L) are polynomials in the lag operator; the disturbance et is assumed to be i.i.d. normal with zero mean and variance σe2. Expected inflation is proxied by past inflation (πte=Σi=1Tαiπti). We impose the sum of the weights on lagged inflation terms to be equal to one to ensure dynamic homogeneity—i.e. no relationship between inflation and real variables in the long-run—which is necessary for the NAIRU to be an economically meaningful concept. The preferred specification uses one lag for inflation. NAIRUt is the natural unemployment rate at period t and is modeled as an unobserved stochastic process assumed to follow a random walk and estimated using the Kalman filter:

NAIRUt=NAIRUt1+ϵt

The disturbance term ϵt is assumed to be i.i.d. normal with zero mean and variance σϵ2 As it is common in the literature, σϵ2 is constrained to be a fraction of σe2 (see Gordon 1997). After comparing the model likelihood from using alternative values, this ratio is set at 0.2.

The potential labor force is given by the working age population and the potential participation rate.3 To estimate the potential participation rate, we use a cyclical-adjustment equation that relies on the gap between the unemployment rate and the NAIRU as an indicator of when actual participation equals potential and that allows for potential participation to grow at a constant rate over one or more historical periods.4 The cyclical-adjustment equation results from combining the following equations:

pt=pt*+θ(utNAIRUt)+ϵtpt*=γ(T1,T2,,TN)

where pt and pt* are actual and potential participation levels; θ is an estimated coefficient on the sensitivity of participation to the unemployment gap; γ is a vector of N coefficients and Ti is a trend variable that takes zero values up to the break point in period i. Historical values for the potential participation rate are calculated as the fitted values from the regression with the unemployment rate constrained to equal the NAIRU at each period. Only one trend break, in 2000, is assumed for the final model specification.

  • Capital factor and TFP – While in principle the capital input does not need to be cyclically adjusted—as the unadjusted level already represents its potential contribution to growth—cyclical variations in the rate at which it is used get reflected in the TFP series if capital is not adjusted. Variations in utilization rates were particularly large in Latvia around the recent crisis. Some studies use data on capacity utilization to infer the potential level of TFP.5 In this study, a simpler approach is taken. The capital factor is assumed equal to the capital stock in the economy, computed according to the perpetual inventory approach (with depreciation assumed to be 8 percent per year), multiplied by the ratio of capacity utilization.6 Potential capital is obtained by imposing capacity utilization at its long-run average. The Solow residual after correcting the capital stock by the degree of capacity utilization is assumed to represent the structural level of efficiency in the economy and is taken as the benchmark potential TFP. Arguably, the resulting TPF series might still include cyclical elements such as variable utilization not captured by the capacity utilization rate used (e.g. as we are not correcting the labor factor by variations in the intensive margin). As a robustness exercise we construct an alternative potential TFP series by HP filtering the benchmark series (with a low smoothing parameter of 6.25). Estimates from both approaches differ mainly in magnitude of the pre-crisis output gap estimate (Figure 3). As the process to obtain the alternative potential TFP series embeds no economic information and might be smoothing out non-cyclical variations, the unfiltered series is used for the benchmark model specification.

uA01fig01

Capacity Utilization

(Capacity utilization, industry, in percent)

Citation: IMF Staff Country Reports 2013, 029; 10.5089/9781475578522.002.A001

Sources: Haver; and IMF staff calculations.

References

  • Benes, J., K. Clinton, R. Garcia-Saltos, M. Johnson, D. Laxton, P. Manchev and T. Matheson, 2010, “Estimating Potential Output with a Multivariate Filter,IMF Working Paper 10/285 (Washington: International Monetary Fund)..

    • Search Google Scholar
    • Export Citation
  • Blanchard, Olivier J. and Danny Quah, 1989, “The Dynamic Effects of Aggregate Demand and Supply Disturbances,American Economic Review, Vol. 79, No. 4, pp. 65573.

    • Search Google Scholar
    • Export Citation
  • Boone, Laurence, Claude Giorno, Mara Meacci, Dave Rae, Pete Richardson and Dave Turner, 2001, “Estimating the structural rate of unemployment for the OECD countries,OECD Economic Studies, vol. 2001(2), pages 171216.

    • Search Google Scholar
    • Export Citation
  • Congressional Budget Office (CBO), 2001, “CBO’s Method for Estimating Potential Output: An Update”.

  • D’Auria, Francesca, Cécile Denis, Karel Havik, Kieran McMorrow, Christophe Planas, Rafal Raciborski, Werner Röger, and Alessandro Rossi, 2010, “The production function methodology for calculating potential growth rates and output gaps”, European Economy, Economic Paper No. 420, European Commission.

    • Search Google Scholar
    • Export Citation
  • Gordon, Robert J., 1997, “The Time-Varying NAIRU and Its Implications for Economic Policy,Journal of Economic Perspectives, American Economic Association, vol. 11(1), pp. 1132.

    • Search Google Scholar
    • Export Citation
  • Ravn, Morten O. and Harald Uhlig, 2002, “On Adjusting the HP-Filter for the Frequency of Observations”, The Review of Economics and Statistics, 84(2): pp. 37138.

    • Search Google Scholar
    • Export Citation
1

Prepared by Bertrand Gruss.

2

We estimate a joint model of quarterly real GDP and unemployment and use a Blanchard and Quah (1989)-type decomposition, relying on long-run economic restrictions, to identify supply and demand shocks. The output gap is computed as the accumulation of demand shocks (to anchor the level of the output gap we assume the 2004 gap is equal to the PF method’s gap estimate). We use quarterly data since 1996 and four lags as suggested by lag length selection criteria.

3

The PF variant used here bears similarities with the one used by the U.S. Congress and Budget Office as described in CBO (2001)

4

Also, the estimate from using a constant-NAIRU model is almost 12 percent.

1

Hours worked was not used in this specifications due to data concerns.

2

An alternative specification using headline inflation and other supply shock control variables (e.g. changes in energy and food prices) was also used and the results were not significantly different.

3

For the purpose of this analysis, labor statistics up to 2011 have been estimated by staff by extrapolating the correction introduced in 2011 due to Census data for the period in between census years (2000–11).

4

This approach is similar to the one used by the Congress and Budget Office to estimate potential output in the U.S.; see CBO (2001).

5

For example, a new method being used by the European Commission to estimate trend TFP relies on the Kalman filter and data on capacity utilization to disentangle cyclical and structural variations in TFP; see D’Auria et al. (2010).

6

The level of capacity utilization is proxied by the capacity utilization in manufacturing from managers’ answers to business and consumer surveys published by the National Statistical Office.

References

  • AIKNC [Higher Education Quality Assessment Center], 2012, http://www.aiknc.lv/lv/list.php

  • AIKOS, 2012, “Institutions of Education and Science,” http://www.aikos.smm.lt/aikos/institutions.htm

  • AIP [Higher Education Council], 2012, “Eiropas Sociālā fonda projekts “Augstākās izglītības studiju programmu izvērtēšana un priekšlikumi kvalitātes paaugstināšanai”,” http://www.aip.lv/ESF_projekts_publ_29.htm

    • Search Google Scholar
    • Export Citation
  • Analītisko pētījumu un stratēģiju laboratorija, 2007, “Profesionālās izglītības prestižs Latvijā”, study commissioned by the Ministry of Education and Science.

    • Search Google Scholar
    • Export Citation
  • Auers, Daunis, Toms Rostoks, and Kenneth Smith, 2007, “Flipping burgers or flipping pages? Student employment and academic attainment in post-Soviet Latvia,Communist and Post-Communist Studies, vol. 40(4): pp.47791.

    • Search Google Scholar
    • Export Citation
  • Barkbu, Bergljot, Jesmin Rahman, and Rodrigo O. Valdés, 2012, “Fostering Growth in Europe Now”, Staff Discussion Note No. 12/07, Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • BICG [Baltic Institute of Corporate Governance], 2010, Baltic Guidance on the Corporate Governance of Government-Owned Enterprises, http://www.corporategovernance.lt/en/news/17

    • Search Google Scholar
    • Export Citation
  • BICG [Baltic Institute of Corporate Governance], 2012, Governance of State-Owned Enterprises in the Baltic States, http://www.corporategovernance.lt/en/news/41

    • Search Google Scholar
    • Export Citation
  • Chapman, Bruce, and Peter Tulip, 2008, “International Dimensions in the Financing of Higher Education”, Australian National University, Discussion Paper No. 574.

    • Search Google Scholar
    • Export Citation
  • CEDEFOP [European Centre for the Development of Vocational Training], 2012a, http://www.cedefop.europa.eu/EN/statistics-and-indicators/initial-vocational-education-and-training.aspx

  • CEDEFOP [European Centre for the Development of Vocational Training], 2012b, “From education to working life,” Publications Office of the European Union, Luxembourg: CEDEFOP.

    • Search Google Scholar
    • Export Citation
  • CEPEJ [European Commission for the Efficiency of Justice], 2012, European Judicial Systems – Edition 2012 – Efficiency and Quality of Justice, CEPEJ Studies No. 18, Council of Europe Publishing.

    • Search Google Scholar
    • Export Citation
  • Council of the European Union, 2012, “Council Recommendation on the National Reform Programme 2012 of Latvia and delivering a Council opinion on the Convergence Programme of Latvia, 2012–15,http://ec.europa.eu/europe2020/europe-2020-in-your-country/latvija/index_en.htm

    • Search Google Scholar
    • Export Citation
  • Diena, 2012a, “Šadurskis: Ķīļa reformas no Latvijas izskalotu gaišākos prātus,http://www.diena.lv/diena-tv/dienas-rits/sadurskis-kila-reformas-no-latvijas-izskalotugaisakos-pratus-13944902

    • Search Google Scholar
    • Export Citation
  • Diena, 2012b, “Vidusskolas piemaksā, lai skolēni neietu uz arodskolām,http://www.diena.lv/latvija/zinas/vidusskolas-piemaksa-lai-skoleni-neietu-uz-arodskolam-13965408

    • Search Google Scholar
    • Export Citation
  • Dombrovsky, Vyacheslav, 2010, “The sorry state of higher education (and research),http://politika.lv/article/the-sorry-state-of-higher-education-and-research

    • Search Google Scholar
    • Export Citation
  • Dombrovsky, Vyacheslav, 2011, “Financing Higher Education: A Model for Reform”, Stratēģiskās analīzes komisija (Riga).

  • Erbenova, Michaela, Yan Liu, and Magnus Saxegaard, 2011, “Corporate and Household Debt Distress in Latvia: Strengthening the Incentives for a Market-Based Approach to Debt Resolution,IMF Working Paper WP/11/85, (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Estonian Ministry of Education and Research, 2009, “Recent developments in higher education governance in Estonia,Public presentation, http://www.slideserve.com/santa/recent-developments-in-higher-education-governance-in-estonia-annika-tina-deputy-head-he-department-estonian-ministr

    • Search Google Scholar
    • Export Citation
  • European Commission, 2012, “EU BOP Assistance to Latvia – First Review Under Post-Programme Surveillance,http://ec.europa.eu/economy_finance/eu_borrower/balance_of_payments/pdf/lv_efc_note_pps_mission_en.pdf

    • Search Google Scholar
    • Export Citation
  • FICIL [Foreign Investors’ Council in Latvia], 2010, “Position Paper on Higher Education in Latvia,http://ficil.lv/index.php/documents/positionpapers/2010/

    • Search Google Scholar
    • Export Citation
  • FICIL [Foreign Investors’ Council in Latvia], 2011, “Position Paper on Capital Markets Development,http://ficil.lv/index.php/documents/positionpapers/2011/

    • Search Google Scholar
    • Export Citation
  • FICIL [Foreign Investors’ Council in Latvia], 2012a, “Position Paper on Proposals for Facilitating Efficiency of Justice,http://ficil.lv/index.php/documents/positionpapers/2012/

    • Search Google Scholar
    • Export Citation
  • FICIL [Foreign Investors’ Council in Latvia], 2012b, “Position Paper on Proposals for the Improvement of the Insolvency Process.

  • FICIL [Foreign Investors’ Council in Latvia], 2012c, “Position Paper on Competitiveness of State-Owned Enterprises.

  • Giudice, Gabriele, Gatis Eglitis, and Christian Weise, 2012, “EU Balance-of-Payments assistance for Latvia: Foundations of success”, Occasional Paper No. 120, European Commission: Economic and Financial Affairs, November.

    • Search Google Scholar
    • Export Citation
  • Ir, 2012a, “Valdība atbalsta augstskolu akreditācijas noteikumus,http://www.ir.lv/2012/9/25/kilis-lidz-gada-beigam-tiks-parvertetas-studiju-programmas

    • Search Google Scholar
    • Export Citation
  • Ir, 2012b, “Ķīlis šoruden rosinās palielināt svešvalodu lietošanu augstskolu studijās”, http://www.ir.lv/2012/9/21/kilis-soruden-rosinas-izmanas-svesvalodu-lietosanaspalielinasanai-augstskolu-studijas

    • Search Google Scholar
    • Export Citation
  • Lawin, 2009, Arbitration in the Baltics, mimeo http://www.lawin.com/files/arbitration_in_the_baltics_2009.pdf

  • Ministry of Economics, 2012, Report on the Economic Development of Latvia, June, Riga.

  • MoES [Ministry of Education and Science], (2009), Koncepcija “Profesionālās izglītības pievilcības paaugstināšana un sociālo partneru līdzdalība profesionālās izglītības kvalitātes nodrošināšana” (informatīvā daļa) [Concept paper “Increasing the attractiveness of the vocational education and raising the participation of the social partners in ensuring the quality of vocational education”]

    • Search Google Scholar
    • Export Citation
  • MoES, 2010a, “Pārskats par Latvijas augstāko izglītību 2010.gadā (galvenie statistikas dati)[Report on Higher Education in Latvia for 2010 (Main Statistical Data)], http://izm.izm.gov.lv/registri-statistika/statistika-augstaka/6818.html

    • Search Google Scholar
    • Export Citation
  • MoES, 2010b, “Pasākumu plāns nepieciešamajām reformām augstākajā izglītībā un zinātnē 2010. – 2012.gadam,http://izm.izm.gov.lv/upload_file/Izglitiba/Augstaka_izglitiba/IZMPlans_280610_AIZref.pdf

    • Search Google Scholar
    • Export Citation
  • MoES, 2011a, “Pārskats par Latvijas augstāko izglītību 2011.gadā (galvenie statistikas dati)[Report on Higher Education in Latvia for 2011 (Main Statistical Data)], http://izm.izm.gov.lv/registri-statistika/statistika-augstaka/8190.html

    • Search Google Scholar
    • Export Citation
  • MoES, 2011b, “Pēc diskusijas ar studentiem Roberts Ķīlis pauž apņēmību sākt aktīvu reformu izskaidrošanu,press release, http://izm.izm.gov.lv/aktualitates/informacijamedijiem/8102.html

    • Search Google Scholar
    • Export Citation
  • MoES, 2012, “Aktualitātes un jaunumi izglītības jomā 2012./2013. mācību gadā,http://izm.izm.gov.lv/aktualitates2012-2013.html

    • Search Google Scholar
    • Export Citation
  • OECD, 2009, Privatisation in the 21st Century:Recent Experiences of OECD Countries—Report on Good Practices, (Paris: OECD).

  • OECD, 2011, Corporate Governance of State-Owned Enterprises: Change and Reform in OECD Countries since 2005, (Paris: OECD).

  • Stockholm School of Economics in Riga, 2012, Latvian Competitiveness Report 2011, http://www.sseriga.edu/en/research/competitiveness-report/

    • Search Google Scholar
    • Export Citation
  • State Employment Agency, 2008, “Darba devējiem sniegto pakalpojumu attīstības stratēģija 2008.–2010. gadam,www.nva.gov.lv/docs/11_488dbc78d14155.57205601.ppt

    • Search Google Scholar
    • Export Citation
  • UNCITRAL, 2005, Legislative Guide on Insolvency Law, (New York: United Nations).

  • Wölfl, Anita, Isabelle Wanner, Oliver Röhn, and Giuseppe Nicoletti, 2010, “Product Market Regulation: Extending the Analysis Beyond OECD Countries”, OECD Economics Department Working Papers, No. 799, OECD Publishing. http://dx.doi.org/10.1787/5km68g3d1xzn-en

    • Search Google Scholar
    • Export Citation
  • World Economic Forum, 2012, Global Competitiveness Report 2012–13, http://www.weforum.org/reports

1

Prepared by Agnese Bukovska and David Moore.

2

Some of the Doing Business subindicators seem more reliable than others. Latvia’s ranking for “getting credit” (4th out of 185) seems too favorable given the extent of deleveraging. And a caveat to Latvia’s unfavorable ranking on “protecting investors” is that Finland has the same ranking (70th out of 185) on this subindicator—which is hard to square with other indicators for Finland showing a particularly strong business environment.

3

However, the NAP proposals on regional development—while consistent with earlier long-range planning documents—are now proving controversial. The NAP envisages focusing resources on 9 larger cities and 21 development centers: the trade-off is that smaller municipalities would miss out, but resources would no longer be spread too thinly to be effective.

5

For a more complete overview of program issues from the IMF perspective, please see the upcoming Ex-Post Evaluation of Exceptional Access Under the 2008 Stand-By Arrangement. For the European Commission’s perspective, see Giudice et al (2012).

6

See for example, IMF Country Report No. 12/171, the staff report for the 2010 Article IV Consultation.

7

See the LCR, and Lawin (2009).

8

For further information see Erbenova et al. (2011). The IMF’s Legal Department provided technical assistance to the MoJ in the design of amendments to the insolvency framework.

9

See, for example, Ir magazine, May 19, 2011.

10

Under section 57(2) of the Law on Insolvency, a legal person is liable to insolvency proceedings if its unpaid debt exceeds LVL 3,000, and if three weeks after a warning from its creditor, “has not paid its debt or raised justified objections to the claim.”

Under section 67 of the Law, an administrator has rights including to sell the property of the debtor (67–1), and to liquidate branches or representation offices of the debtor (67–2).

The Law does not provide for an appeal against a court decision to launch insolvency proceedings and appoint an administrator. UNCITRAL (2005) recommends allowing the debtor a right to appeal, though without an appeal granting suspensive effect (so as to deter frivolous appeals).

11

See KNAB (2012).

12

See the Staff Report for the Fifth Review Under the Stand-By Arrangement (IMF Country Report No. 12/31), Box 1.

13

Baltic Institute of Corporate Governance, Baltic Guidance on the Governance of Government-owned Enterprises, June 2010. Available at: http://www.corporategovernance.lt/en/news/17

16

HEIs that provide higher professional education are colleges. These institutions also provide vocational secondary education.

18

According to the Law on Higher Education, the 12-member Council of Higher Education (CHE) includes representatives from 9 educational organizations: the Latvian Academy of Science, the Association of Art Higher Education Institutions, the Latvian Association of Education Managers, the Colleges Association of Latvia, the Council of Rectors, the Latvian Association of University and College Professors, the Education and Science Workers Trade Union, the Latvian Students Association, and a representative of non-state HEIs. The CHE also includes 2 business group representatives and the education minister ex officio.

20

Chapman and Tulip (2008) provide a concise overview of international experience with income-contingent student loans (ICL) and other forms of financing tuition. They caution that the effectiveness of ICL depends on the effectiveness of a country’s income tax system to track graduate incomes and collect debts.

21

This is a main finding of the Latvian Competitiveness Report 2011.

22

Apart from state vocational schools, VEIs are local government-owned or private. Including non-state VEIs, the total number of VEIs was down from 85 in 2009 to 65 in 2011. In 2011, around 88 percent of vocational education students were enrolled in state vocational schools.

Colleges—educational institutions which provide 1st level higher vocational education programs, as well as vocational education and secondary vocational education programs—are considered as HEIs rather than VEIs in the national statistics.

23

Currently each vocational education student receives a state stipend of 7 lats per month and can apply for an additional ESF stipend of 10 to 50 lats per month. This would be in force until the end of 2015.

24

Interview with Skaidrite Abrama, The Baltic Course, July 6, 2012.

Republic of Latvia: Selected Issues
Author: International Monetary Fund. European Dept.