We examine the relationship between the experience of unemployment in the early years of the socio-economic transition in Poland and a number of wellbeing measures about two decades later. The analysis takes advantage of the rich content of data from the Survey of Health, Ageing and Retirement in Europe (SHARE) by matching retrospective information on labour market experiences with outcomes observed in the survey after year 2006. While there is a strong correlation between unemployment and general wellbeing measures such as life satisfaction, depression and subjective assessment of material conditions, the relationship cannot be interpreted as causal. On the other hand, we find that unemployment in the early years of the transition has strong, negative, long-term consequences for income and house ownership. The analysis sheds light on the implications of unemployment and on the nature of job losses in the follow-up of the Polish ‘shock-therapy’.
Next year, the countries of Central and Eastern Europe will celebrate the 30th anniversary of the political breakthrough and the beginning of a major socio-economic transformation which followed. In the Polish case, the ‘shock therapy’ approach to the reform process implemented by the Mazowiecki government, though not without faults, has generally been viewed as the origin of the country’s economic success story. Afterwards, Poland experienced nearly three decades of uninterrupted economic growth and the Polish GDP returned to its pre-reform level already in 1995.
However, discussions of negative implications of the reform package still fuel the academic discourse as well as the political debate. While the majority of the population managed to avoid significant economic difficulties, many families experienced the painful hardship of the transition period in the form of job losses, poverty and exclusion. Given the scale of the socio-economic change, surprisingly little is known about the long-term consequences of individual experiences at that time. In particular, it is unclear if the negative outcomes observed many years after the reforms started can be causally linked to individual experiences in the early 1990s.
This lack of evidence is not unique for Poland and is largely due to unavailability of good individual-level data spanning the time before and after the collapse of communism. Since the transition cannot be lived through again, we shall never know how socio-economic conditions would have looked like under numerous alternative reform scenarios. However, as we show in a recent paper (Myck & Oczkowska, 2018), much can be learnt from the combination of contemporary and retrospective information on the nature of labour market histories during the transition and their relationship to outcomes recorded many years later.
The analysis presented in Myck and Oczkowska (2018) relies on the treatment of the systemic changes in the early 1990s as a major exogenous shock and on differentiating between reasons behind individual experiences of unemployment. We demonstrate that the observed strong correlation between unemployment in the initial years of the transition and a number of subjective wellbeing measures in later life is endogenous, and may reflect unobservable individual characteristics. It seems plausible to argue that these characteristics were the reasons behind the recorded job losses once the economy was liberalised and firms could fire their least productive employees.
Work histories in the SHARE dataset
The analysis is based on individual-level data from the Polish part of the Survey of Health, Ageing and Retirement in Europe (SHARE). SHARE is a multidisciplinary biennial panel survey focusing on individuals aged 50 years and over. Since the start of the project in year 2004 seven waves of data have been collected, and the survey was conducted in Poland in waves 2, 3, 4, 6 and 7. While the standard waves of the survey focus on contemporary conditions of respondents such as health, economic conditions, labour market activity and social networks, in wave 3 (the so-called SHARE-Life), participants were asked about their life histories including their family history, mobility and labour market experiences. The detailed labour market histories recorded in SHARE-Life allow us to identify transition-related job losses, which can be matched with current information on several measures of material conditions and wellbeing for the same individuals.
In Figure 1 we present labour market profiles since 1988 of those in the sample who were working prior to the start of the reform process.
Figure 1. Labour market status 1988 – 2008 conditional on working in 1988 in Poland
Source: Myck and Oczkowska, 2018.
The figure shows that along with rapidly increasing unemployment rates, the degree of inactivity of the Polish population grew substantially in the two decades following the transition. This data confirms that in the follow-up of the ‘shock-therapy’ reforms many individuals faced unemployment, while others, especially among older groups of employees, used several other labour market exit options, such as retirement or disability.
Analysing long-term consequences of economic shocks
To examine the role of unemployment experiences in the initial years of the transition for outcomes observed a few decades later, we use data from waves 2, 3 and 4 of the SHARE study. The analysis focuses on two groups of later-life outcomes – objective measures of material conditions such as household income, real assets and house ownership, and subjective indicators of wellbeing such as life satisfaction, depression or reporting difficulties in making ends meet.
We are able to control for an extensive set of individual characteristics which are usually unobservable to the researcher, through a complex set of background variables available in SHARE. These include respondents’ childhood conditions, parental background as well as health and labour market experience prior to 1988. With regard to the experience of unemployment we differentiate the instances of unemployment between the initial (1989-1991) and later (1992-1995) period of the transition to examine the potential differential implications of the rapid pace of the reforms in the early 1990s. Most importantly though, the data allows us to distinguish between different reasons behind job losses and we can separately examine the relationship with plant/office closures and other reasons for unemployment. Following other examples in the literature (Farber, 2011; Jacobson et al., 1993), we argue that plant closures can be treated as reasons for exogenous job separations. This in turn allows us on the one hand, to give a causal interpretation to the estimated coefficients, and on the other, to interpret those on other reasons for unemployment in the light of the causal relations.
Effects of unemployment experience on later-life outcomes
We find that experiencing unemployment due to plant/office closure between 1989 and 1991 is associated with almost a 30 percent lower level of household income and a lower probability of house ownership of about 10 percentage points (pp) some two decades afterwards. There is also a strong relationship between unemployment in the early years of the transition and wellbeing measures two decades later – individuals who experienced unemployment in the first three years of the transition have a 14 pp. higher likelihood of reporting great difficulties in making ends meet, a 10 pp. lower probability of high life satisfaction and a 11 pp. higher likelihood of depression. However, since these relations do not hold for unemployment due to plant closures, they cannot be treated as causal. The results are therefore most likely driven by unobserved factors which simultaneously determine the lower level of outcomes two decades after the ‘shock-therapy’ reforms, and the likelihood of experiencing unemployment in the early 1990s.
In this policy brief we outline recent results on long-term implications of labour market developments in the early years of the economic transition in Poland. The analysis is based on a combination of contemporary and retrospective data from the SHARE survey, and focuses on the associations between the experience of unemployment in the initial years of the transition in Poland and a number of outcomes measured about two decades later. Using plant/office closures as exogenous sources of job separations during the early 1990s, we find a strong and statistically significant, negative, long-term effect on income and home ownership, which can be treated as causal.
We also find strong negative associations between unemployment for other reasons than plant / office closures and a number of subjective measures of wellbeing. This relationship however, does not hold for the exogenous reasons for job losses, which suggests an important role of unobservable factors that lead to unemployment and at the same time are responsible for the lower level of outcomes in later life. This is consistent with the labour market reality of central planning characterised by labour hoarding and maintaining employment regardless of workers’ productivity. When the economic reality changed in 1989, the least productive individuals were the first to be fired, and as our analysis shows, these are also the individuals with lower subjective levels of wellbeing two decades later. We confirm thus that those who lost their jobs in the early 1990s have lower measures of the subjective wellbeing outcomes, although the latter cannot be identified as specific consequences of unemployment in the first years of transition.
- Farber, H. (2011). “Job loss in the great recession: historical perspective from the Displaced Workers Survey, 1984-2010”, NBER Working Paper No. 17040, National Bureau of Economic Research.
- Jacobson, L., LaLonde, R. and Sullivan, D. (1993). “Earnings losses of displaced workers”, American Economic Review, 83, pp. 685–709.
- Myck, M. and Oczkowska, M. (2018). “Shocked by therapy? Unemployment in the first years of the socio-economic transition in Poland and its long-term consequences”, Economics of Transition, 26(4), pp. 695-724.
The authors gratefully acknowledge the support of the Polish National Science Centre through project no. 2015/17/B/HS4/01018. For the full list of acknowledgements see Myck and Oczkowska (2018).
Twenty years after transition began and the merits of “shock therapy” were argued the most hotly, (former) transition countries are hard hit by global shocks originating in Western market economies. Although discussions now focus on troubles in Western developed countries, countries in Eastern Europe and the CIS were particularly hard hit in 2008/09. This should not come as a surprise given their pre-crisis vulnerabilities. As transition countries opened up to trade and capital flows—like other countries that want to reap the benefits of the global economy—they also became subject to the shocks that hit open economies. The very high current account deficits and/or reliance on commodity exports prior to the crisis exposed many countries in this region to two of the shocks that have been most costly to emerging markets and developing countries in the past, namely sudden stops in capital flows and terms of trade shocks. However, the lesson from the crisis should not be that opening up is bad in general, but that shock absorbers should be put in place to minimize the damage market-based “shock therapy” can do.
Shock Therapy and Transition
It is now twenty years since the failed 1991 coup that led to the breakup of the Soviet Union and started the transition from plan to market in (then) communist countries. A few years earlier, in 1989, the Stockholm Institute of East European Economies was set up at the Stockholm School of Economics. Its first director was Anders Åslund who was a strong proponent of shock therapy (see Åslund 1992). The basic idea was that a rapid transition from plan to market through changing institutions, privatizations and other liberal reforms would minimize the cost of transition. There are still different views on the merits of shock therapy as a reform strategy, but this brief will not address this debate.
In the wake of the academic debate of shock therapy, the broader research agenda was put under the heading “transition economics”, which became a new field in economics. This also lead Erik Berglöf, the new director of the Stockholm Institute of East European Economies, to change the institute’s name to the Stockholm Institute of Transition Economics, or SITE for short, the name we still use today. The economics of transition was also the title of the fifth Nobel symposium in Economics (Berglöf and Roland, 2007).
The use of the label “transition economics” may see a revival in the aftermath of the Arab spring. However, the economic issues that now face the (former) transition countries are much the same issues that emerging markets, developing countries and also advanced countries around the world grapple with. This is not least true when we look at what has happened in the current crisis.
Below this brief will argue that the shock therapy ex-communist countries were subject to in the early 90’s has been changed to the shock “therapy” open market economies around the world have been facing for a very long time. And just as there were heated debates on what the right therapy was then, the world is now debating what the “therapy” for the current shocks should be.
Output Drops Around the World
Significant drops in output have not only been observed in countries transitioning from plan to market but have occurred with regularity throughout modern history in countries around the world. The figure from Becker and Mauro (2006) shows the frequency of countries that entered into a major output loss event—defined as a cumulative loss of at least 5 percent of initial GDP in an event that last at least two years—during the 20th century. In the after-war period, a relative modest 5 to 10 percent of countries have entered into a period of significant output loss. However, in the great depression, almost half of the countries for which data is available entered into a period of significant loss of output.
Since the methodology used in Becker and Mauro follows countries until they return to pre-crisis levels of GDP, it is too early to provide a full account of what the number would be in the current crisis that started in 2008. Nevertheless, it is straightforward to compute how many countries have experienced output losses in 2008/09 (which is the first criteria to satisfy in the Becker and Mauro measure) and this number is close to fifty percent, on par with the great depression. This is not to say that the total output loss will be as high as in the great depression, but it clearly tells the story that this is the worst global crisis we have seen since then in terms of how many countries have been affected. The share of countries affected at the onset of this crisis varied greatly in different parts of the world and at different levels of development. The most surprising statistic from this crisis is that 90 percent of advanced economies experienced an output drop whereas “only” 40 percent of emerging market countries did. The regional differences between emerging markets are also striking; 85 percent of countries in Central and Eastern Europe (CEE) and more than half of CIS countries saw output decline, far above other regions.
The Becker and Mauro paper also looks at the correlates of major output drops and focuses on a number of domestic and external macro, financial and political shocks as triggers of output collapses. A large dataset of shocks and output drops is used to compute the frequency of the different shocks; the likelihood that a particular shock leads to an output collapse (as defined above); and the output loss associated with such event. These numbers are then used to calculate how much different types of shocks have cost in terms of lost output for emerging markets and developing countries. The scale in the chart shows how much it would be worth in terms of GDP per annum to avoid a certain shock.
For emerging market countries, the worst shock has been sudden stops in capital flows that cost almost a percent of GDP per year. Unless countries have high levels of foreign exchange reserves, sudden stops in capital flows mean that (often large) current account deficits have to contract and even become surpluses quickly because there is no external financing available for a deficit. This, in turn, affects domestic production and demand and can have a serious effect on output. Add to this the financial uncertainty that is often associated with sudden stops and it makes it the number one shock to worry about for emerging market countries integrated in the global economy and financial markets.
Less developed countries are in several cases dependent on concentrated commodity exports to generate foreign exchange. This makes this group of countries more vulnerable to terms-of-trade shocks. The estimates of how costly these shocks are range from around half a percent of GDP per year, as is shown in the chart, to around 2.5 percent of GDP if a more extensive sample including very long-lasting output events is used.
Other shocks like currency, political and debt crisis have also been associated with significant losses in output, but tend to occur less frequently, which makes the cost per year smaller.
The 2008/09 Crisis
As was mentioned previously, not enough time has passed since the start of the crisis in 2008 to use the methodology in Becker and Mauro to compute cumulative output losses since many countries are still below their pre-crisis GDP levels. However, projected losses can be computed by using the IMF’s World Economic Outlook forecasts of GDP growth rates. If we then rank the countries according to worst output performance, CEE and CIS countries dominate the “top-ten” list with seven entries. Latvia at the top of the list is projected to lose a cumulative 40 plus percent of pre-crisis GDP during the 11 years it is projected to take the country to return to the GDP level it enjoyed in 2007. The other Baltic countries, Ukraine and Armenia have also been hit particularly hard in this crisis. Russia and Romania are close behind three advanced countries that had to seek IMF and EU assistance to deal with the crisis; Ireland, Iceland and Greece.
The forecasts used for these calculations are constantly being revised and the ranking of countries based on actual outcomes will certainly look different years from now. We can only hope that the current forecasts are too pessimistic although right now many revisions go in the wrong direction.
Could we have predicted what countries were hit in this crisis based on history? Based on the Becker and Mauro (2006) paper: Yes and no. “No” regarding the fact that overall, many advanced countries were hit this time. The paper found that in the past, the frequency of output collapses have decreased with the income level of countries, contrary to what has been the case this time. “Yes” for the fact that CEE and CIS countries were hit significantly more that other emerging markets since they were particularly vulnerable to the sudden stops and terms-of-trade shocks that the paper showed often are associated with severe output losses.
The signs of vulnerability in CEE and CIS countries were easy to see, but warnings ignored; the Baltic countries and Romania had double-digit current account deficits that where to a large part financed by external debt. For example, current account deficits in Latvia passed 20 percent of GDP in some years, far beyond the 4-5 percent deficits some Asian countries had prior to the Asian crisis in 1997. Ukraine also had large current account deficits and concentrated (metal) exports on top of that, exposing the country to both sudden stops and terms-of-trade shocks.
Russia’s dependency on energy and mineral exports was also well known, with 80 percent of export revenues coming from this source. However, the pre-crisis boom in oil and mineral prices had made Russian policy makers think this was a strength and not a vulnerability. On top of that, the strong external position of the government and central bank obscured the external financial vulnerabilities that had built up in the private sector. With the sharp decline in oil prices in the crisis, Russia was hit both by a terms-of-trade shock and a sudden stop in capital flows to the private sector.
There were of course individual countries in other regions that were showing vulnerabilities and were hit in the crisis, but the countries in the CEE and CIS region stood out as particularly vulnerable to the shocks that hit emerging market and developing countries in the past. The IMF’s Global Financial Stability Report from October 2008 summarizes these vulnerabilities in its Table 1.5 on macro and financial indicators for emerging markets very well. The shaded boxes that indicate potential problems in the table completely dominated the picture for CEE and CIS countries whereas other regions looked significantly less vulnerable. In short, history told us what shocks matter and the vulnerability indicators clearly showed where the shocks would do most damage.
What are the policy lessons from this? Countries will always be subject to different types of shocks, and the question is how these shocks can best be absorbed to minimize the economic costs. In other words, what is the “therapy” needed to deal with these shocks? A key challenge is to find the policies and tools that strike the right balance between crisis prevention and high sustainable growth. Isolation to protect against external shocks is certainly not the answer.
Early warning systems (EWS) that identify vulnerabilities ahead of time and give policy makers time to reduce these vulnerabilities and thus avoid crisis is of course what everyone dreams of. The IMF and others have worked on various EWS models since the Mexican and Asian crises with mixed success (see Berg, Borensztein and Pattillo, 2005). With this crisis, the G20 and other groups of policy makers have made new calls for developing EWS, at times seemingly unaware of past efforts and the limited success in this area.
However, at the IMF the more formalized or mechanical EWS models are complemented with both bi-lateral and multilateral surveillance with the bulk of the findings made publicly available and communicated to relevant policy makers. These surveillance efforts contain much more information than more limited EWS models and also come with policy recommendation on how to deal with potential vulnerabilities.
There are of course limitations also with the surveillance by the IMF and other international and domestic organizations. First of all, they have to get it right and at the right time. This is far from trivial, not least because of our limited understanding of the links between the financial sector and the real economy. And even when the analysis gets it “right” in the sense of identifying vulnerabilities, it may take a long time before vulnerabilities translate into real problems and during this time, the analysis and recommended policies may seem misguided.
This is linked to the second major issue; to make relevant policy makers (politicians) listen to and take action on the advice. There is a reason Reinhart and Rogoff called their book “This time is different” since this is often the response to warnings from the IMF and others that suggest a party has been going on for too long and the punch bowl needs to be taken away.
Given the limitations of early warnings and surveillance more generally, there remains a strong need to reduce vulnerabilities in a systematic manner and develop tools to deal with the crises that were not prevented. This will require both domestic measures and a strong commitment to international cooperation. The latter part is of course extremely important right now in order to find appropriate solutions (read debt resolutions) to the problems cross-border banking and capital flows have created. It will also be key in setting up the rules for the future: what capital requirements should banks have; (how) should financial transactions or institutions be taxed; how can cross-border supervision be made more efficient; what type of crisis resolution mechanisms should be put in place both at the international and regional levels; etc, etc. Unless there is broad international agreement on these issues they will do little to address the weaknesses that were at the heart of this crisis.
On the domestic side, the usual IMF recommendation of creating a stable macroeconomic environment—with fiscal room to maneuver and a monetary policy that leads to stable prices—is always going to be part of a countries ability to absorb shocks. For countries that are integrated in international financial markets, exchange rate flexibility and a reasonable level of international reserves seem to be advisable. Jeanne and Rancière (2008) analyze optimal foreign exchange reserves for countries that are subject to sudden stops. Becker (1999) instead looked at accumulation of government assets as part of an optimal public debt and asset management strategy in a world with bailouts of the private sector which seems particularly relevant today.
The macro side should of course be combined with strong domestic supervision of the financial sector; structural policies that lead to sustainable growth in a competitive global environment; and strategies in commodity exporters to reduce the vulnerabilities associated with a narrow export base.
Although advanced countries get most of the attention in the international financial press today, emerging market and developing countries should not think that this is a new world were the shocks of the past do not matter to them. They do, so better get ready for “shock therapy” the market way while there still is time.
- Becker, T., (1999), “Public debt management and bailouts”, IMF WP 99/103.
- Becker, T. and P. Mauro, (2006), “Output drops and the shocks that matter”, IMF WP 06/172.
- Berg, A., E. Borensztein, and C. Pattillo, (2005), “Assessing Early Warning Systems: How Have They Worked in Practice?”, IMF Staff Papers, 52:3, 462-502.
- Berglöf, E. and G. Roland (2007), The economics of transition—The fifth Nobel symposium in economics, Stockholm Institute of Transition Economics (SITE) and Palgrave Macmillan.
- IMF (2008), Global Financial Stability Report, “Financial stress and deleveraging”, October, Table 1.5 p.46.
- Jeanne, O. and R. Rancière, (2008), “The optimal level of international reserves for emerging market countries: A new formula and some applications”, CEPR Discussion Papers 6723.
- Reinhart, C. and K. Rogoff, (2009), “This time is different: eight centuries of financial folly”, Princeton University Press.
- Åslund, A., (1992) “Post-Communist Economic Revolutions: How Big a Bang?”, Center for Strategic & International Studies, Washington D.C..
Disclaimer: Opinions expressed in policy briefs and other publications are those of the authors; they do not necessarily reflect those of the FREE Network and its research institutes.