Chapter 1. Trends and challenges in the Latvian labour market

This chapter provides an overview of recent economic and labour market developments in Latvia, draws on a range of data sources to analyse current unemployment from several angles and identifies vulnerable groups of jobseekers in the Latvian labour market.

The statistical data for Israel are supplied by and under the responsibility of the relevant Israeli authorities. The use of such data by the OECD is without prejudice to the status of the Golan Heights, East Jerusalem and Israeli settlements in the West Bank under the terms of international law.

Recent labour market trends: from crisis to stagnating recovery

Over the last two decades, developments on Latvia’s labour market were shaped by a long economic boom before 2008, an especially severe recession in the years 2008-2010 and a strong recovery (Figure 1.1). The boom before 2008 was fuelled by the prospect and initial effects of Latvia’s accession to the European Union in 2004, which led to an abundant supply of credit at low interest rates and expectations of rapid income gains (Blanchard, Griffiths and Gruss, 2013[1]). Wages and house prices rose quickly, encouraging consumption and thereby reinforcing an economic upward spiral (OECD, 2015[2]).

Figure 1.1. Boom, bust and recovery in the Baltic states, 1998-2018

When access to credit became more restricted in 2007, the falling house prices undermined consumption – a development that was greatly aggravated by the sudden stop of lending from foreign banks at the beginning of the financial crisis in September 2008 (Åslund and Dombrovskis, 2011[3]). While the Latvian GDP had been growing at an average rate of 10% annually between 2000 and 2007, it declined by about one-quarter between 2007 and 2010 (OECD, 2015[2]). Coinciding with the bust after Latvia’s boom, the global financial crisis thus had an especially strong impact on Latvia, much stronger than in most other OECD countries but very similar to the impact on Estonia and Lithuania, where a comparable boom period had preceded the crisis (Figure 1.1).

In all three Baltic states, the severe recession was soon followed by a strong recovery (Figure 1.1). Latvia’s economic growth since 2011 has been one of the highest in Europe (OECD, 2015[2]) and climbed to 5% in the first quarter of 2018. The economic situation is expected to improve further as Latvia benefits from robust domestic consumption, from growing export markets in both the euro area and the Russian Federation, as well as from resuming investment of EU funds (OECD, 2017[4]). While negative effects of recent sanctions against the Russian Federation and of Russian counter-sanctions may have been very limited (Oja, 2015[5]), geopolitical tensions with Russia remain perhaps the largest risk for the Latvian economy (OECD, 2017[4]).

The recession strongly reduced employment and increased unemployment

Efforts to respond to the crisis unfolding between 2007 and 2009 could not prevent large impacts on wages and employment. Latvia’s currency at the time, the lat, had been pegged to the euro since 2005. Maintaining a stable exchange rate and meeting the Maastricht criteria to eventually join the European currency union were political priorities (Purfield and Rosenberg, 2010[6]) and Latvia could indeed adopt the euro in January 2014. To the large shock of the global financial crisis, Latvia therefore did not react with an external devaluation, but with an internal devaluation that requires a reduction of domestic prices – including wages – to regain competitiveness on exports markets. Because prices and wages adjust only slowly, the short-run adjustment of the labour market fell on employment.

By consequence, employment and unemployment in Latvia exhibited dramatic changes (Figure 1.2). The employment rate for the population aged 15-74 fell from 63% in Q1 2008 to 51% in Q1 2010 (Panel A). While employment rates also declined rapidly in Estonia and Lithuania (by close to 10 and 7 percentage points, respectively), the fall in Latvia was significantly larger. Unemployment rates rose steeply over this period, by 13-15 percentage points in all three Baltic states (Panel B). Coming from a slightly higher initial level, only Latvia’s unemployment rate reached 20% of the total labour force. These large movements contrast with very moderate changes in the average employment and unemployment rates for the OECD area and underline the severity of the crisis impact on the Baltic states.

Figure 1.2. Employment trends in the Baltic states, 2007-2017 (projections to 2019)

The rise of unemployment in Latvia by up to 15 percentage points was substantially higher than in most other OECD countries that were heavily affected by the crisis, including Portugal and Ireland where unemployment rose by up to 9 and 10 percentage points, respectively (Figure 1.3). The rise in Latvia was only exceeded by the increases observed in Spain and Greece of up to 18 and 20 percentage points, respectively. By the end of 2015, however, the gap in Latvia had fallen below 5 percentage points, approaching the gaps in Portugal and Ireland and underlining the relative speed of Latvia’s recovery. The gap had remained very large in Spain and Greece (12 and 16 percentage points, respectively).

Figure 1.3. Change of unemployment in OECD countries over the financial crisis, 2007-2015

During the recovery, unemployment halved but stagnated above pre-crisis levels

Already in 2010, Latvia’s unemployment rate started falling and the employment rate started rising again (Figure 1.2, Panel B). Between Q1 2010 and Q1 2015, the unemployment rate halved, falling from above 20% to just below 10%. However, the decline has since slowed down, and the unemployment rate is projected to stay around 8% in 2018/2019. This unemployment rate is still somewhat higher than before the financial crisis. It also corresponds to one and a half times the average unemployment rate for the OECD area, and this difference is expected to remain throughout 2018/2019.

The evolution of Latvia’s employment rate shares some of the same features. However, its recovery has continued since 2010 at roughly the same pace (Figure 1.2, Panel A). By early 2018, the employment rate reached the highest pre-crisis levels (63%), and it is expected to approach 65% in 2019. Already matching the OECD average, Latvia’s employment rate could thus exceed it significantly for the first time since 2008.

The gradual slowdown of the improvement in Latvia’s unemployment rate suggests that its recovery after the financial crisis may be approaching its end. The unemployment rate already appears to have stabilised at a level well above the OECD average, and the projections shown in Figure 1.2 highlight the risk that the unemployment rate remains by and large unchanged despite robust economic growth and expanding employment. At this stage, further reductions of unemployment might require policies that address structural unemployment rather than cyclical unemployment linked to crisis effects.

The participation rate in Latvia has recently climbed beyond pre-crisis levels (Figure 1.4). After a drop in 2009 largely offset increases observed in 2007/2008, the participation rate had recovered by 2012. During 2015 and 2017, it rose to significantly higher levels. At close to 78% in early 2018, Latvia’s participation rate substantially exceeds the OECD average (72% in 2017) as well as those in other OECD countries that suffered heavily from the financial crisis, such as Greece (68%), Spain and Portugal (both 75%). Only few OECD countries exhibited substantially higher participation rates than Latvia: Iceland (88%), Switzerland (84%), Sweden (83%), New Zealand (81%) and the Netherlands (80%).

Figure 1.4. Evolution of labour market participation in Latvia, 2007-2018

However, Figure 1.4 also indicates that the rise in the participation rate is partly driven by demographic developments. Latvia’s population of working age (15-64 years) fell significantly between 2007 and 2018, due to population ageing and high levels of emigration (OECD, 2016[8]). The active population of working-age also declined over this period, but not as much as the total population of working age. The narrowing gap between them translates into a rising participation rate.

After a large decline, strong growth of real wages has resumed

As part of Latvia’s internal devaluation, wages came under heavy downward pressure (Figure 1.5). After real wages had grown by 10% annually in the period Q1 2000 to Q4 2007, they declined by 11% annually in the period Q4 2007 to Q1 2009. In the following years, real wages stabilised and returned to strong growth between 2012 and 2015, but without fully offsetting the earlier decline. These swings in real wages were the most extreme observed among OECD countries, and considerably larger than in other OECD countries that were strongly affected by the financial crisis, such as Greece, Ireland, or Spain. Comparable real wage changes were only observed in Lithuania and, albeit to a much smaller extent, in Estonia, which again highlights parallels between developments in the Baltic states. Latvia was unique among the Baltic states to also experience declining nominal wages – a fall by 6% annually in the period Q4 2007 to Q1 2009 (OECD, 2015[2]). In the public sector, nominal wages were cut by up to 30% (Raudla and Kattel, 2013[9]).

Figure 1.5. Real wage changes in OECD countries, 2007-2015

The downward pressure on wages may have been alleviated to some extent by considerable rises in Latvia’s legal minimum wage. During the crisis period, the minimum wage in nominal terms rose from LVL 2 045 (the national currency before the euro) in 2007 to LVL 2 726 in 2008, then to LVL 3 067 in 2009/2010 and LVL 3 408 in 2011-2013. In relative terms, the minimum wage represented 37% of the median wage of full-time workers in 2007, rose to 47% in 2009 and reached 51% in 2011. After receding slightly in the following years, the minimum wage again equalled 51% of the median wage in 2016 (Figure 1.6). While this level did not stand out among OECD countries, the increase of Latvia’s minimum wage in relative terms was stronger than in any other OECD country: between 2007 and 2016, it increased by 39%. The only comparable increases in OECD countries were observed in Lithuania (31%) and Poland (37%). In nominal terms, Latvia’s minimum wage more than doubled over this period.

Figure 1.6. Evolution of national minimum wages in OECD countries, 2007-2016

As stressed in OECD (2016[10]), the rapid rise of Latvia’s minimum wage may have had a strong impact because comparatively many workers earned wages in 2010 that would fall below the minimum wage in subsequent years. Especially for labour market entrants without significant work experience and for low-skilled workers, the minimum wage may be too high compared to their productivity, so that their chances of finding employment are undermined. However, as Zasova (2011[11]) points out, this reasoning does not apply where low wages are recorded by official statistics but workers informally receive higher wages in practice, so-called envelope wages. Based on survey data, Žukauskas and Schneider (2016[12]) estimate that almost four-fifths (79%) of those who worked at least partly in the shadow economy in 2015 received envelope wages, more than the corresponding shares for Lithuania (70%), Estonia (66%) and Poland (64%). The remainder are unregistered self-employed, who accounted for 21% in those working in Latvia’s shadow economy compared to 38% in Lithuania, 29% in Estonia and 13% in Poland.

The shadow economy has likely shrunk during the recovery

Using surveys among entrepreneurs, Putniņš and Sauka (2018[13]) can combine estimates of envelope wages and wages paid to undeclared employees with estimates of undeclared business income. They find that Latvia’s shadow economy is sizeable but in recent years considerably smaller than during the crisis period: according to their results, the shadow economy accounted for around 37% of GDP in 2009/2010, then began decreasing to 30% in 2011 and has since been fluctuating between 20% and 24% (Figure 1.7). Throughout this period, the estimated shadow economy in Latvia was larger than those in Estonia and Lithuania (18% in 2017). Undeclared employment in Latvia was estimated at 15% of total employment in 2009/10, fell to 12% in 2011 and then continued declining, reaching 7% in 2017.

Figure 1.7. Estimates of the shadow economy in the Baltic States, 2009-2017

One interpretation of the estimates in Figure 1.7 is that some economic activity shifted to the shadow economy during 2008-2010 in an effort to reduce costs by saving on taxes and social security contributions (Vanags, 2012[14]). This shift may have reversed when the situation in the formal economy improved during the recovery. Because the shadow economy may have provided many with an alternative to the formal economy, the dramatic changes of Latvia’s GDP, employment and unemployment in the wake of the financial crisis should not be taken fully at face value. Production that shifted to the shadow economy would result in a loss of GDP in the formal economy, and likewise for employment – among those counted as unemployed, some likely worked informally. While the observed changes in GDP, employment and unemployment may therefore overestimate the true adjustment, the shadow economy could in all likelihood only dampen the adverse effects and not nearly offset them.

Some sectors thrive but others have not recovered

While total employment in Latvia’s formal economy has recovered in recent years, it is still below pre-crisis levels: 895 000 persons were employed in 2017, according to data from Latvia’s Central Statistical Bureau. This level significantly exceeded total employment in 2010 (851 000) and at the beginning of the recovery in 2011 (862 000). However, total employment reached the same level as in 2017 already in 2013 and has since tended to fluctuate rather than grow. Given the decline of the working-age population, it seems unlikely that total employment can return to the substantially higher level at the end of the boom in 2008 (1.06 million) in the next years.

In addition, the recovery of employment has been uneven across sectors: employment in 2017 was in some sectors higher than in 2008, but in most sectors it was still much lower than in 2008 (Figure 1.8). In the sector comprising finance, insurance, real estate, business administration and support service as well as professional, scientific and technical activities, employment was 29% higher in 2017 than in 2008. Employment gains also occurred in health and social work (plus 15%), partly driven by high demand in these occupations from public employers in response to legal requirements. In all other sectors shown in Figure 1.8, employment in 2017 was still lower than in 2008, despite often substantial employment growth in the period 2011-2017. Employment in construction has not nearly recovered from the end of Latvia’s housing boom, falling by one-half between 2008 and 2017. Another large decline was observed in agriculture, where employment was still decreasing in the period 2011-2017 and fell by 27% between 2008 and 2017. Over these years, employment in public administration and in manufacturing (including mining, energy, water and waste management) fell about as strongly (minus 25% and 22%, respectively), but returned to substantial growth of around 6% in the period 2011-2017.

Figure 1.8. Evolution of employment in Latvia by sector, 2008-2017

This uneven development across sectors likely complicates the reduction of unemployment. The skills demanded for positions in the two thriving sectors will often not match the skills of unemployed jobseekers whose experience or training falls into different sector. In so far as sectors were concentrated in certain Latvian regions, those who lost their jobs in these sectors might often be unable to take up new jobs unless they move to another region. The slow recovery of sectors also means that unemployed persons with skills specific to these sectors might not have used their skills in years, which makes them less employable.

In the next few years, the uneven development of employment in sectors might well continue. According to recent CEDEFOP projections, a substantial increase in employment is only expected in business and market-oriented services (plus 2% between 2017 and 2020). Limited increases are expected for manufacturing, trade and transport as well as construction. Decreasing employment is forecast for non-marketed services and especially for the primary sector and utilities (minus 6%).

Not all unemployed have benefited from the recovery

Since employment in many sectors has not recovered, not all of those who became unemployed in the wake of the economic crisis have eventually found formal employment in Latvia again. Instead, many may have retreated from the labour market and moved to inactivity, the shadow economy or another country. Figure 1.9 depicts an attempt to estimate the flow of persons who left unemployment but did not take up formal employment in Latvia. This flow is estimated as the residual change of unemployment levels after accounting for total inflows and for outflows to employment. The underlying data on labour market flows were constructed by Fadejeva and Opmane (2016[15]) from changes in individual labour market status observed in Latvia’s Labour Force Survey.

Figure 1.9. Flows into and out of unemployment in Latvia, 2008-2014

The results in Figure 1.9 suggest that substantial numbers of persons who left unemployment did not take up formal employment in Latvia, even during the recovery. At least 160 000 exits from unemployment during 2008-2014 (28% of all outflows) were not part of the flow into formal employment in Latvia. The true figures are most likely higher because total inflows are underestimated, which leads to an underestimation of residual outflows. To some extent, these outflows may reverse if substantial numbers of emigrants return in the context of Brexit, so that the initial effect might be to raise unemployment in Latvia (Box 1.1).

Discouraged workers remain comparatively frequent

Labour force survey data allow identifying those inactive persons of working age who do not seek work (anymore) because they believe that they will not find any. Persons with this self-reported, subjective belief are known as discouraged workers and are often regarded as hidden unemployed – they would be counted as unemployed if they actively sought work. The number of discouraged workers in Latvia returned to pre-crisis levels of about 20 000 in 2015 and 2016, after nearly tripling during the crisis years and reaching 47 000 in 2010 (Figure 1.10, Panel A). The number of long-term discouraged workers – those who have been without employment for more than 12 months – followed a very similar trajectory. Their share among all discouraged workers remained above 70% throughout the recovery and temporarily climbed to 85% in 2015.

Despite the recent decreases, the number of discouraged workers remains high in Latvia in comparison to other European OECD countries (Figure 1.10, Panel B). They represented 5% of Latvia’s non-employed working-age population in 2016, exceeding the corresponding shares in all but two European OECD countries: Portugal (8%) and Italy (11%). Several countries that were heavily affected by the financial crisis exhibited much lower shares of discouraged workers, such as Spain (3%), Ireland (2%) and Greece (1%). In Estonia and Lithuania, the corresponding shares amounted to 3% and below 2% in the period 2012-2016, respectively, after increasing only mildly in the wake of the financial crisis. For labour market policy in Latvia, these findings highlight the particular challenge to reach persons who have retreated from the labour market and lack motivation to search for jobs, also in a context of improving labour market conditions. According to evidence from the Latvian Labour Force Survey for 2012-2016, three-quarters of discouraged workers wish having a job (see Chapter 2). To better understand the barriers faced by discouraged workers, their individual characteristics are examined towards the end of this chapter.

The increase in labour demand is accelerating

During the recovery, the prospects of finding a job have improved due to rising labour demand: the vacancy rate in Latvia rose from 1.4% of all non-agricultural positions in the first quarter of 2012 to 2.5% in the first quarter of 2018, largely exceeding pre-crisis levels of around 2% (Figure 1.12). While Latvia’s vacancy rate was initially on a par with that of Estonia and the EU average, the subsequent increase was significantly stronger in Latvia. However, much of this increase occurred only during the first quarter of 2018: previously, the vacancy rate in Latvia had gradually increased to 1.9% in the last quarter of 2017.

Figure 1.10. Discouraged workers in Latvia and other OECD countries

Most sectors of the Latvian economy exhibited vacancy rates from about 1% to about 3% of all positions in the first quarter of 2018. Especially high values were observed in wholesale and retail trade (3.1%) accommodation and food services (3.4%), construction (3.8%), and public administration (5.1%). In some cases, high vacancy rates might not only reflect high labour demand but also working conditions that make these jobs unattractive. At the lower end, only 0.6% of all positions in education were recorded as vacant.

Box 1.1. Labour market outcomes of returning Latvian emigrants

Just over 3 000 Latvian citizens returned from abroad in 2017 (Siliņa-Osmane, 2018[16]). This helps explain why return migrants contribute at least initially Evidence on the labour market outcomes of return migrants in recent years suggest that they contribute, at least initially, to unemployment in Latvia (Figure 1.11). At 28% in the first year after return, their unemployment rate in Latvia is much higher than it had been abroad one year earlier, and it is twice as high as the average unemployment rate in Latvia (Panel A).

The very high unemployment rate – and a correspondingly low employment rate – one year after return highlight the challenge for returnees of integrating in the Latvian labour market. In fact, a substantial share of returnees of working age (15-64) change labour market status: while 14% change from inactivity or unemployment abroad into employment in Latvia, 11% change from employment abroad into unemployment in Latvia (Figure 1.11, Panel B). The latter group likely includes some who return to Latvia after losing their job abroad, but also many who choose to return e.g. for family reasons without having a job lined up in Latvia. According to a survey of return migrants in 2016 (Hazans, 2016[17]), the primary motives for return are related to family and friends in Latvia (indicated by 60%), while few had returned due to an attractive job offer (6%).

That returning emigrants at least initially face problems to find employment in their home country is not unique to Latvia. Martin and Radu (2012[18]) document that emigrants returning to Central and Eastern European countries are more often inactive than the general population and, in the first year after return, also more often unemployed. However, returning emigrants who find employment can earn substantially higher wages than their peers who did not emigrate, according to the results in Hazans (2008[19]) for Latvia, in Martin and Radu (2012[18]) for nine Central and Eastern European countries including Latvia and in Tverdostup and Masso (2016[20]) for Estonia.

Reintegration policies can support return migrants in their efforts to settle in Latvia again. A report on reintegration policies at the local level found that 25% of municipalities, especially often those in regions Latgale and Vidzeme, consider reintegration of return migrants a very relevant policy issue (Žabko et al., 2017[21]). Ensuring that children of return migrants can enter school or kindergarten – which often requires learning Latvian – appears to be the main challenge. Supporting these pupils is therefore one of the most frequent reintegration policies (indicated by 14% of surveyed municipalities), alongside assistance with finding accommodation (11%) and the provision of information targeted at return migrants (8%). In 2018, the Ministry of Environmental Protection and Regional Development launched a pilot project aimed at coordinating reintegration efforts across regions (Siliņa-Osmane, 2018[16]). As part of this initiative, financial support of up to EUR 9 000 is available to return migrants for starting a business.

Future return migration to Latvia could be affected when the United Kingdom exits from the EU (“Brexit”). The United Kingdom hosts more emigrants from Latvia than any other OECD country (OECD, 2015[22]). By 2016, a total number of 113 000 citizens of Latvia were residing in the United Kingdom, based on data from the Office for National Statistics (Salt, 2016[23]). This number declined slightly in comparison to 2015 (117 000) and 2014 (121 000). The number of employed among them fell more strongly over these years: from 73 000 in 2014 and 68 000 in 2015 to 64 000 in 2016.

Figure 1.11. Labour market outcomes of emigrants returning to Latvia, 2012-2016

The status of EU citizens currently allows Latvians to move freely to the United Kingdom, take residence and find work without the need for residence or work permits. This may change profoundly: depending on the outcome of negotiations between the United Kingdom and the European Union, Latvian citizens’ residence and access to the labour market may be subject to conditions, annual quotas or limited durations.

If such limitations on the rights of Latvians in the United Kingdom do not only apply to newly arriving emigrants but also to those who are already resident, outmigration of Latvians from the United Kingdom will likely increase. Survey results compiled by M. Hazans and reported in OECD (2016[8]) indicate that economic factors and search for better quality of life where the main drivers of Latvian emigration. Therefore, restrictions on labour market access in the United Kingdom could prompt many Latvians to move elsewhere, notably to other countries in the EU. Given the recovery of the Latvian economy in recent years and the family ties of Latvian emigrants, some of them might also return to Latvia.

Figure 1.12. Labour demand in European OECD countries, 2012 and 2018

Shortages of skilled labour contrast with rather low wage differentiation

In some parts of the Latvian economy, labour shortages have developed over recent years. According to the business and consumer surveys of the European Commission’s Directorate General for Economic and Financial Affairs, one in five employers in Latvia indicated that labour shortages hold back their production in the last quarter of 2017 (seasonally adjusted). While this share did not stand out compared to other European OECD countries, it marked the highest level for Latvia since 2008. From 2013 to 2016, the share in Latvia had fluctuated around only one in ten employers.

The labour shortages appear to arise in particular for skilled labour, but also for service workers. For skills in terms of specific knowledge, Figure 1.13 shows the most recent available indicators for shortages and surpluses in Latvia. Overall, these indicators suggest significant shortages of many types of skills. By far the most severe shortages were observed for skills in computers and electronics, followed by roughly equal shortages in mathematical and engineering skills, service skills and administrative skills. The largest surpluses occurred for skills in building and construction and skills in production and processing.

Figure 1.14 provides further indications that labour demand in Latvia is strong relative to labour supply, so that many employers have to offer attractive working conditions to fill their vacancies. Most employers appear to offer permanent contracts to new hires: only 35% of new hires in 2016 received a fixed-term contract, which was the third-lowest proportion among European OECD countries. It is also rather rare that employees in Latvia are overqualified for their job. At 13%, Latvia’s over-qualification rate in 2015 fell into the lower half of European OECD countries. While Finland, the Czech Republic and Poland exhibited significantly lower over-qualification rates (7%-10%), Portugal’s over-qualification rate was almost double that in Latvia (25%).

Figure 1.13. Skill needs in the Latvian economy, 2015

Figure 1.14. Contract duration for new hires and over-qualification rates in European OECD countries, 2015/2016

At the same time, the average wage premium associated with skills in terms of formal degrees remains rather low in Latvia: employees’ earnings with a tertiary education are 143% of earnings with upper secondary education (Figure 1.15). Most OECD countries exhibit a higher premium for a tertiary education. Similarly, the premium for upper secondary education is rather low in Latvia: earnings with lower education still reach 88% of earnings with upper secondary education. However, in the case of Latvia, figures refer to wages net of income tax, which tend to be less differentiated than gross wages.

Figure 1.15. The wage premium for skills in OECD countries, 2015 or latest available year

Macro and micro-level characteristics of unemployment in Latvia

This section focuses on specific aspects that, at the macro level, help characterise current unemployment in Latvia in comparison to other OECD countries and, at the micro level, identify demographic and socio-economic groups who experience unemployment particularly often or more often for long durations.

Remaining unemployment may be largely structural

For the period 2009-2012, Figure 1.16, Panel A shows that the observed unemployment rate in Latvia far exceeded the estimated natural rate of unemployment, i.e. the unemployment rate that prevails when cyclical forces on the labour market are in equilibrium. This suggests that the especially high levels of unemployment in these years were largely cyclical, reflecting the end of Latvia’s boom and the impact of the financial crisis. Since 2013, however, the observed unemployment rate has been close to the natural rate of unemployment, which suggests that unemployment in recent years can primarily be explained by structural factors.

Latvia’s natural rate of unemployment in 2017 is estimated at 9%, a high level in comparison to other OECD countries (Figure 1.16, Panel B). Similar natural rates of unemployment are estimated for France, the Slovak Republic, Italy and Portugal, while estimates are substantially higher only for Greece and Spain (15-16%). Estimates are lower for Lithuania (8%) and Estonia (6%), suggesting that Latvia is most affected by structural unemployment among the Baltic States.

Figure 1.16. Structural component of the Latvian unemployment rate

The finding of comparatively high structural unemployment in Latvia aligns with various estimates of Latvia’s natural rate of unemployment provided in the literature. According to estimates by Krasnopjorovs (2015[25]), Latvia’s natural rate of unemployment stood at 11% in 2014 and was very close to observed unemployment rates in 2013 and 2014 after large cyclical divergences in earlier years. Ebeke and Everaert (2014[26]) estimate the natural rate of unemployment to be about 13% at the end of 2013, with an average level of 12.3% over the period Q1 2002 to Q4 2014. Based on data from 1990 to 2013 and from 1996 to 2008, respectively, Blanchard, Griffiths and Gruss (2013[1]) and Meļihovs and Zasova (2009[27]) obtain estimates of 13.3%, and 11%. While Anosova et al. (2013[28]) do not estimate a natural rate of unemployment, they conclude that the rise of unemployment in the wake of the crisis and the subsequent recovery were predominantly cyclical, in line with Figure 1.16, Panel A.

Two-fifths of the unemployed have been unemployed for more than one year

At 3.3% in 2017, the rate of long-term unemployment (unemployment for 12 months or more) in Latvia is close to twice the OECD average (Figure 1.17). Only seven OECD countries exhibited higher long-term unemployment rates than Latvia, while rates were lower notably for Estonia (2.0%) and Lithuania (2.7%). At the same time, Latvia’s long-term unemployment rate was one of the lowest among the countries that were heavily affected by the financial crisis. Long-term unemployed made up about two-fifths (38%) of all unemployed in Latvia in 2017. While this share substantially exceeded the OECD average (31%), it roughly matched the shares in the Czech Republic, Japan and Switzerland, and was close to the shares in the other Baltic States. Significantly higher shares were again observed in other countries that were heavily affected by the crisis.

Figure 1.17. Long-term unemployment in OECD countries, 2017

In the wake of the financial crisis, long-term unemployment in Latvia had increased strongly, albeit somewhat later than total unemployment – large numbers of newly unemployed persons became long-term unemployed only over time (Figure 1.18, Panel A). During the recovery, long-term unemployment decreased steadily. At 32 000 in 2017, the level of long-term unemployment was about one-third of the level in 2010 and below pre-crisis levels. The share of long-term unemployment in total unemployment has, however, remained above pre-crisis levels: despite a declining tendency, it only fell to 38% in 2017, compared with 20-30% in the period 2007-2009.

Figure 1.18. Unemployment duration in Latvia

Persons with especially long unemployment durations persistently make up a substantial share of total unemployment in Latvia (Figure 1.18, Panel B). Since the beginning of 2011, the share of unemployment durations of four years or more has fluctuated between 15% and 21% of total unemployment, without a clear tendency to decrease or increase. The highest shares were recorded as recently as 2015. In 2017, durations of four years or more accounted for 15% of the unemployed, below but still close to the average for the period since 2011 (17%). The sustained decline in the overall share of long-term unemployment, which fell below 40% in 2017, was most strongly driven by the declining share of unemployment durations from two to four years. Long durations of unemployment can have serious adverse effects on jobseekers’ ability to find suitable employment: valuable skills and know-how may be lost when they are not used, professional contacts and networks can dissolve over time, and in some cases jobless persons might adopt detrimental habits that can permanently undermine their employability, such as regular alcohol abuse.

Staying jobless for a long duration can be demoralising: the unemployed jobseekers might reduce the intensity of their job search if they experience it to be pointless, to the extent that they give up altogether and join the ranks of discouraged workers (Krueger and Mueller, 2011[29]). This can broadly be confirmed for Latvia, using data from the European Labour Force Survey for 2007-2016: up to one month after the end of their last job, only 7% of persons not in employment are discouraged workers, i.e. they consider it pointless to engage in job search. The share rises to around 9% for 1-5 months after the last job, and further to around 10% for 6 months to 4 years. Among those who last held a job more than four years ago, 15% are discouraged workers.

Interregional differences in unemployment are large

Across Latvian regions, the incidence of unemployment, youth unemployment and long-term unemployment varies widely (Figure 1.19). In Riga and Pieriga, unemployment rates were comparatively low in 2016 (between 6% and 8%), while rates of 12% and 18% were recorded in Zemgale and Latgale, respectively. In the age group 15-24, unemployment rates significantly exceeded 20% in Vidzeme and Latgale, but were about half as high in Riga and Pieriga. Similarly, the share of long-term unemployed was about 1.5 times as high in the former two regions as in the latter two.

Data from the Central Statistical Bureau show that employment levels have recovered in Riga, Pieriga and Zemgale between 2010 and 2016 (but still remain below pre-crisis levels), while hardly any employment gains have been observed in Latgale, Kurzeme and Vidzeme. The divergence is linked to the performance of sectors in these regions. Employment in construction declined strongly and permanently in all Latvian regions. Employment in trade, accommodation and food-related services also declined strongly in most Latvian regions. In addition, strongly falling employment in industry has affected Riga, Kurzeme and Latgale, but not the other regions. Kurzeme and Latgale further had to deal with a strong and permanent employment decline in agriculture and fishing, as well as in public administration.

With different employment prospects across regions, raising the mobility of unemployed persons becomes a policy objective, so that they can go where the jobs are. Chapter 4 of this Review evaluates a programme intended to encourage taking up jobs offers that would require moving or commuting over significant distances. Alternatively, unemployed persons can be supported in setting up businesses in their current location, and Chapter 4 also offers an assessment of a programme that fosters entrepreneurship among unemployed persons through coaching and the provision of grants.

Figure 1.19. Unemployment in Latvian regions, 2016

Unemployment rates are higher for youth and older men

Unemployment rates by age group indicate that those aged 15-24 face a substantially higher risk of unemployment, with a rate of around 15% in 2017 (Figure 1.20, Panel A). While this applies to men and women alike, only men exhibit relatively high unemployment rates at ages 45 and above. By contrast, women in these age groups have especially low unemployment rates (at most 6%). While 45% of all male unemployed are long-term unemployed, this applies to only 37% of unemployed women (Panel B).

The difference in unemployment rates at ages 45 and above partly reflects higher educational attainment among women in these age groups, notably a larger share with a high education level. According to further results for 2016 based on the same data (the European Labour Force Survey), 41% of the female labour force aged 45-54 have a high education level, compared with 21% of the male labour force. This holds similarly in the age group 55-64: while 38% of the female labour force are highly educated, only 24% of the male labour force reach this level. As unemployment rates are lower for high education levels (see below), women’s higher educational attainment translates into lower unemployment rates than observed for men.

Figure 1.20. Employment and unemployment in Latvia by sex and age group, 2016/2017

However, in both age groups, women also exhibit lower unemployment rates than men with the same level of education. For example, the unemployment rate for women aged 45-54 with a low education level was 15% in 2016, compared with 24% for low-educated men aged 45-54. These differences likely result from several factors. Jobless women appear more likely to retreat from the labour market: in the age group 45-54, 32% of women not in employment are unemployed rather than inactive, compared with a figure of 42% for men. In the age group 55-64, this disparity is especially large: only 12% of women not in employment are unemployed rather than inactive, but still 21% of men. The differences in unemployment rates could also reflect differences in the reservation wages of the unemployed: especially in the age group 45-54, reservation wages of unemployed men are substantially higher than for women with the same level of education Figure 1.21. These reservation wages might be based on previous wages and specific experience rather than current labour market prospects, so that they become an obstacle to finding employment in new job roles, sectors or even occupations where this experience is partly irrelevant. Such changes to a very different work environment may, however, be necessary for those who are affected by unemployment after the age of 45.

Figure 1.21. Gender gap in reservation wages of registered unemployed aged 45-64 by educational attainment, 2012-2017

In the average for the age groups 45-54 and 55-64, the low unemployment rates of women cancel with the high unemployment rates of men. As a result, only the average unemployment rate of those aged 15-24 stands out, reaching 17% in 2016 while averages for other age groups remained below 10% (Figure 1.20 Panel C). The age groups 15-24 and 55-64 have in common that they exhibit substantially lower employment rates (33% and 61%, respectively) than other age groups (about 80%).

The share of long-term unemployed is about one-half for all age groups from 35, while this share is significantly lower for younger age groups (Figure 1.20, Panel C). At an early stage of the working life, it may be easier to avoid long-term unemployment by changing between occupations or regions than at a stage when one is committed to a particular occupation and region. To some extent, however, this result also reflects that many young labour market entrants cannot possibly be long-term unemployed because they have not yet participated in the labour market for more than a year.

While high in comparison to other age groups, the unemployment rate of those aged 15-24 does not appear particularly elevated in comparison to other OECD countries (Figure 1.22). In 2017, other countries that were heavily affected by the financial crisis exhibited substantially higher youth unemployment rates than in Latvia. At 12%, the OECD average was nevertheless substantially lower than the youth unemployment rate in Latvia. With regards to older men, however, the situation in Latvia stands out: the unemployment rate of men aged 55-64 (10%) was one of the highest among OECD countries and was exceeded only in Spain (15%) and Greece (17%). At 2.6 percentage points, the gap to the unemployment rate of women aged 55-64 was higher in Latvia than in all other OECD countries except Estonia and Turkey. These findings suggest that older men are an especially vulnerable group on Latvia’s labour market.

Figure 1.22. Unemployment rates of youth and older workers in OECD countries, 2017

In Latvian active labour market policy, unemployed persons aged 55 and above are considered a vulnerable group. They are therefore eligible for a programme of subsidised employment: for up to one year, direct subsidies to an employer decrease the costs of employing a person in the programme. The employment of persons with disabilities can be subsidised for up to two years. Chapter 5 in this Review focuses on the situation of some of the most vulnerable groups among the unemployed and evaluates the impact of the subsidised employment programme.

Many long-term unemployed lack education or work experience

Employment and unemployment rates in Latvia differ widely between levels of education (Figure 1.23, Panel A). Between those with a high education level and those with a low education level, employment rates differed by 50 percentage points in 2016. While the unemployment rate of those with a high education level stood at 4%, it reached 21% for those with a low education level. The shares of the long-term unemployed were around 45% for unemployed with low and medium education levels, but 27% for unemployed with a high education level.

A high education level thus seems to be associated with a substantially lower risk of unemployment and long-term unemployment. Employment in Latvia has for years exhibited a trend towards greater employment of the highly-educated and smaller employment of the low-educated. Further analyses using the same data as in Figure 1.23 show that the employment share of highly-educated persons increased from 24% in 2007 to 37% in 2016. Over the same period, the employment share of low-educated persons declined from 13% to 8%, and that of medium-educated persons from 63% to 55%.

Upgrading the skills of unemployed persons is therefore a key concern for active labour market policy in Latvia, as in virtually all OECD countries. In terms of participants, training is the most used active labour market programme in Latvia, and a number of training programmes allow unemployed persons to acquire a formal qualification. Chapter 3 of this Review offers a detailed evaluation of the main training programmes for unemployed persons in Latvia. It examines effects that only materialise after several years and gives special attention to the role of training in reducing long-term unemployment.

In principle, low formal education levels may be mitigated by work experience and skills learnt on the job. Yet the lack of prior work experience is especially wide-spread among unemployed with low education levels, concerning 30% of them in 2016 but only 20% and 23% of unemployed with high and medium education levels, respectively (Figure 1.23, Panel A). The lack of prior work experience also implies a high risk of long-term unemployment, which affects close to three-quarters of unemployed without experience, in both educational groups (Figure 1.23, Panel B). For unemployed persons with work experience, the shares of the long-term unemployed are far lower: 31% for those with high and medium education levels and 38% for those with a low education level.

The risk of long-term unemployment also depends on how the previous job ended (Figure 1.23, Panel B). The share of long-term unemployment is relatively high among persons who stopped working because of illness or disability and those who were dismissed (37% and 40%, respectively). It remains comparatively low in those cases where a limited-term employment contract ended (29%) and for those who chose to leave a previous job in order to look after children or due to other family responsibilities (25%). In line with this result, further analyses do not find a strong role for demographic characteristics of households: both unemployment rates and the share of long-term unemployment do not appear systematically linked to the presence of young children or elderly in the household.

Discouraged workers are most frequent at medium education levels, where they account for 6% of those not in employment (Figure 1.23, Panel A). The corresponding share among the low-educated with prior work experience approaches 9%, but reaches only 3% among low-skilled without work experience (Figure 1.23, Panel B). This reflects large differences by age groups: further analyses show that the share of discouraged among low-educated persons is above 8% in all age groups except for the age group 15-24 (1%), where many do not yet have work experience. The share among low-educated aged 45-54 stands out, surpassing 16%. The share of discouraged workers is also relatively high among those who were dismissed from their last job or reached the end of a fixed term (8%-9%), but is much lower where the last employment ended for family or health reasons (3%).

The findings in this and the previous section highlight the importance of individual characteristics and employment histories. Evidence from 2007-2010 on persons with persistent labour market difficulties suggests that certain characteristics often occur together, so that several groups can be delimited (Ferré, Immervoll and Sinnott, 2013[30]). The largest such groups in Latvia were identified as single older persons who are unemployed or disabled (22% of persons with persistent labour market difficulties), single young men with a low education level (18%), older unemployed workers who are nevertheless able to work (14%). Three further groups – stay-at home mothers with a small child, low-educated male breadwinners in rural areas and self-employed older men – each accounted for a share of about 10%.

Figure 1.23. Employment, unemployment and discouraged workers in Latvia by educational attainment and employment history, 2016

Over the last years, active labour market policy in Latvia has responded by differentiating measures more strongly between target groups. A host of measures that specifically target young persons aged 15-29 were introduced under the Youth Guarantee. One of the most recently introduced programmes – the special programme for the activation of long-term unemployed – takes a holistic approach towards the individual situation of long-term unemployed persons, who often face multiple barriers to finding employment. More generally, a profiling system introduced in 2013 seeks to ensure that the measures taken by Latvia’s public employment service are well targeted. Chapter 2 of this Review describes current labour market policy in Latvia in comparison to other OECD countries, with a focus on the functioning of the public employment service (the State Employment Agency, SEA) and the delivery of its services and activation measures.

In conclusion, the older men in Latvia, youth and persons with a low level of education are especially often affected by unemployment. The risk that it turns into long-term unemployment is especially high for unemployed without prior work experience, but relatively low for unemployed with a high education level. Living in some rural regions also appears linked to a higher risk of unemployment or long-term unemployment, or both as in the case of Latgale. Although total levels of unemployment and long-term unemployment in Latvia have fallen comparatively rapidly during the recovery from the financial crisis, long-term unemployment still presents a challenge and many sectors have not returned to pre-crisis levels of employment.

References

[28] Anosova, D. et al. (2013), “Structural or cyclical? Unemployment in Latvia Since the 2008-09 Financial Crisis”, Discussion Paper, No. 9525, CEPR.

[3] Åslund, A. and V. Dombrovskis (2011), How Latvia came through the financial crisis, Peterson Institute for International Economics, Washington, D.C.

[1] Blanchard, O., M. Griffiths and B. Gruss (2013), “Boom, bust, recovery: Forensics of the Latvia crisis”, Brookings Papers on Economic Activity, Vol. 2013/2, pp. 325-388, https://doi.org/10.1353/eca.2013.0015.

[26] Ebeke, C. and G. Everaert (2014), “Unemployment and structural unemployment in the Baltics”, IMF Working Papers, No. 2014/153, IMF, Washington, D.C, https://doi.org/10.5089/9781498317207.001.

[15] Fadejeva, L. and I. Opmane (2016), “Internal labour market mobility in 2005–2014 in Latvia: The micro data approach”, Baltic Journal of Economics, Vol. 16/2, pp. 152-174, https://doi.org/10.1080/1406099X.2016.1196872.

[30] Ferré, C., H. Immervoll and E. Sinnott (2013), “Profiling of People with No or Limited Labor-market Attachment”, in Latvia: Who is Unemployed, Inactive of Needy? Assessing Post-crisis Policy Options, World Bank, Washington, D.C., http://www.lm.gov.lv/upload/aktualitates/lv_profiling_270513.pdf.

[17] Hazans, M. (2016), Atgriešanās Latvijā: remigrantu aptaujas rezultāti, University of Latvia Diaspora and Migration Centre, Riga, http://www.diaspora.lu.lv/fileadmin/user_upload/lu_portal/projekti/diaspora/petijumi/Atgriesanas_Latvija_-_petijuma_zinojums_FINAL03.pdf.

[19] Hazans, M. (2008), “Post-enlargement return migrants’ earnings premium: Evidence from Latvia”, University of Latvia, Riga, https://doi.org/10.2139/ssrn.1269728.

[25] Krasnopjorovs, O. (2015), “Natural and cyclical unemployment in Latvia: new insights from the Beveridge curve model”, Discussion Paper, No. 2-2015, Latvijas Banka, Riga, https://www.bank.lv/images/stories/pielikumi/publikacijas/dp_2_2015_en.pdf.

[29] Krueger, A. and A. Mueller (2011), “Job search, emotional well-being, and job finding in a period of mass unemployment: Evidence from high-frequency longitudinal data”, Brookings Papers on Economic Activity, Vol. 2011/1, pp. 1-57, https://doi.org/10.1353/eca.2011.0001.

[18] Martin, R. and D. Radu (2012), “Return migration: The experience of Eastern Europe”, International Migration, Vol. 50/6, pp. 109-128, https://doi.org/10.1111/j.1468-2435.2012.00762.x.

[27] Meļihovs, A. and A. Zasova (2009), “Assessment of the natural rate of unemployment and capacity utilisation in Latvia”, Baltic Journal of Economics, Vol. 9/2, pp. 25-46, https://doi.org/10.1080/1406099X.2009.10840460.

[24] OECD (2017), Education at a Glance 2017: OECD Indicators, OECD Publishing, Paris, https://doi.org/10.1787/eag-2017-en.

[4] OECD (2017), OECD Economic Outlook, Volume 2017 Issue 1, OECD Publishing, Paris, https://doi.org/10.1787/eco_outlook-v2017-1-en.

[10] OECD (2016), “A more productive labour force in Latvia”, in OECD Reviews of Labour Market and Social Policies: Latvia 2016, OECD Publishing, Paris, https://doi.org/10.1787/9789264250505-7-en.

[7] OECD (2016), “Recent labour market developments and the short-term outlook”, in OECD Employment Outlook, OECD Publishing, Paris, https://doi.org/10.1787/empl_outlook-2016-5-en.

[8] OECD (2016), “Responding to the decline of Latvia’s population”, in OECD Reviews of Labour Market and Social Policies: Latvia 2016, OECD Publishing, Paris, https://doi.org/10.1787/9789264250505-6-en.

[22] OECD (2015), “Latvia”, in Connecting with Emigrants, OECD Publishing, Paris, https://doi.org/10.1787/9789264239845-94-en.

[2] OECD (2015), OECD Economic Surveys: Latvia 2015, OECD Publishing, Paris, https://doi.org/10.1787/9789264228467-en.

[5] Oja, K. (2015), “No milk for the bear: The impact on the Baltic states of Russia’s counter-sanctions”, Baltic Journal of Economics, Vol. 15/1, pp. 38-49, https://doi.org/10.1080/1406099X.2015.1072385.

[6] Purfield, C. and C. Rosenberg (2010), “Adjustment under a Currency Peg: Estonia, Latvia and Lithuania during the Global Financial Crisis 2008-09”, IMF Working Paper, No. 10/213, https://doi.org/10.5089/9781455205448.001.

[13] Putniņš, T. and A. Sauka (2018), Shadow economy index for the Baltic countries 2009-2017, SSE Riga, Riga, http://www.sseriga.edu/sites/default/files/2018-07/sseriga_shadow_economy_index_2009-2017.pdf.

[9] Raudla, R. and R. Kattel (2013), “Fiscal stress management during the financial and economic crisis: The case of the Baltic countries”, International Journal of Public Administration, Vol. 36/10, pp. 732-742, https://doi.org/10.1080/01900692.2013.794428.

[23] Salt, J. (2016), International migration and the United Kingdom. Report of the United Kingdom SOPEMI correspondent to the OECD, University College London, London.

[16] Siliņa-Osmane, I. (2018), Latvia. Report 2018 to the OECD Expert Group on Migration, Office of Citizenship and Migration Affairs, Riga.

[20] Tverdostup, M. and J. Masso (2016), “The labour market performance of young return migrants after the crisis in CEE countries: the case of Estonia”, Baltic Journal of Economics, Vol. 16/2, pp. 192-220, https://doi.org/10.1080/1406099X.2016.1233729.

[14] Vanags, A. (2012), “Undeclared work - the Latvian variants”, Mutual Learning Programme 2012 Autumn Peer Reviews, http://ec.europa.eu/social/BlobServlet?docId=10736&langId=en.

[21] Žabko, O. et al. (2017), Pašvaldību politikas instrumenti aizbraukušo iedzīvotāju remigrācijas veicināšanai, Baltic Institute of Social Sciences, Riga, http://petijumi.mk.gov.lv/sites/default/files/title_file/BISS_remigracija_atsk.docx.

[11] Zasova, A. (2011), “Labour market institutions: An obstacle or support to Latvian labour market recovery?”, Baltic Journal of Economics, Vol. 11/1, pp. 5-23, https://doi.org/10.1080/1406099X.2011.10840488.

[12] Žukauskas, V. and F. Schneider (2016), “Micro Based results of shadow labour market in the Baltic States, Poland, Sweden, and Belarus”, Applied Economics: Systematic Research, Vol. 10/2, pp. 117-134, https://doi.org/10.7220/AESR.2335.8742.2016.10.2.7.

Database references

AMECO database (Eurostat), https://ec.europa.eu/info/business-economy-euro/indicators-statistics/economic-databases/macro-economic-database-ameco/ameco-database_en.

BURVIS Database (State Employment Agency).

Business and consumer surveys (DG ECFIN), https://ec.europa.eu/info/business-economy-euro/indicators-statistics/economic-databases/business-and-consumer-surveys_en.

CEDEFOP, http://www.cedefop.europa.eu/en/publications-and-resources/data-visualisations.

Central Statistical Bureau of Latvia, http://www.csb.gov.lv/en.

European Labour Force Survey (Eurostat), http://ec.europa.eu/eurostat/web/lfs/overview.

Eurostat Labour Market Statistics, http://ec.europa.eu/eurostat/web/labour-market.

Latvian Labour Force Survey (CSB), https://www.csb.gov.lv/en/statistics/statistics-by-theme/social-conditions/unemployment/tables/metadata-employment-and-unemployment.

OECD Economic Outlook Database (No. 103), May 2018, http://stats.oecd.org//Index.aspx?QueryId=51396.

OECD Labour Force Statistics Datasets: http://stats.oecd.org//Index.aspx?QueryId=9594, http://stats.oecd.org//Index.aspx?QueryId=9571 and http://stats.oecd.org//Index.aspx?QueryId=9593.

OECD Minimum Wage Database, https://doi.org/10.1787/data-00313-en.

OECD Quarterly national accounts database, https://stats.oecd.org/Index.aspx?DataSetCode=QNA.

OECD Short-Term Labour Market Statistics, https://stats.oecd.org/index.aspx?queryid=35253.

OECD Short-Term Labour Market Statistics, Unemployment Rates by age and gender, http://dotstat.oecd.org//Index.aspx?QueryId=36499.

OECD Skills for Jobs Database, www.oecdskillsforjobsdatabase.org.

End of the section – Back to iLibrary publication page