053 053 This paper is a product of the Poverty and Equity Global Practice Group. It is part of a larger effort by the World Bank to provide open access to its research and contribute to development policy discussions around the world. The authors may be contacted at ccancho@worldbank.org. The Poverty & Equity Global Practice Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. ‒ Poverty & Equity Global Practice Knowledge Management & Learning Team This paper is co-published with the World Bank Policy Research Working Papers. Why so gloomy? Perceptions of Economic Mobility in Europe and Central Asia1 César Cancho, María E. Dávalos and Carolina Sánchez-Páramo Keywords: subjective well-being, relative well-being, economic mobility, inequality, unfairness, trust, Europe and Central Asia. JEL classification: I3, J6, P2 1  This study was prepared as part of the program on economic mobility led by María E. Dávalos (Senior Economist,  Poverty GP, Europe and Central Asia region) and Carolina Sánchez‐Páramo (Practice Manager, Poverty GP, Europe  and Central Asia region), and financed by the TFESSD (P127607). The authors thank Maurizio Bussolo and  participants of the ECA poverty team internal seminar for the comments received.  Contents   1. Growing but Unhappy: Making Sense of the Puzzle ...............................................................................  3  2. Conceptual Framework for the Analysis of Perceptions of Mobility ....................................................... 5  3. Individual and National Level Data ........................................................................................................  8  4. Perceptions of Economic Mobility and Their Correlates  .........................................................................  9  a)  What are the perceptions of economic mobility observed in ECA? ................................................ 9  b)  What is the socio-demographic profile of those with perceptions of downward mobility? ............ 12  c)  What variables are correlated with perceptions? A multivariate analysis ........................................ 14  d)  Are perceptions similar across sub-regions? ..................................................................................  20  5. Conclusions .......................................................................................................................................... 21  References  ................................................................................................................................................ 23  Appendices .............................................................................................................................................. 27  A1. Sample Characteristics ...................................................................................................................  27  A2. Regression methodology ................................................................................................................  33  A3. Robustness checks .........................................................................................................................  38  A4. Perceived Movements by Individual Characteristics: Heads and Other Member ............................ 42  2 | P a g e     1. Growing but Unhappy: Making Sense of the Puzzle Eastern Europe and Central Asia (ECA) experienced considerable economic growth during the 2000s, recovering after a deep recession in the 1990s and in most cases surpassing the levels of income observed before transition. During the 1990s, countries in the region applied market-oriented reforms that changed the organization of their economies. Throughout all this transition period, economic performance was heterogeneous across countries, but most of them experienced a severe contraction in their economies, and consequently, falls in their standards of living. In the 2000s, in contrast, the landscape was dramatically different. Growing continuously since the early 2000s, by mid-2000s the vast majority of countries had already surpassed their levels before transition, with some of them even getting close to Western European standards (Figure 1). On average, GDP per capita was 30% higher in 2006 than in 1990 (World Bank 2013a), driven mostly by New EU Member States2 and Turkey. In line with the observed economic growth, social indicators indicated a reduction in the incidence and severity of poverty in the region from 1990 to the 2000s. Using the regional poverty line of $2.5/day, the poverty headcount in Europe and Central Asia fell from around 12% in 1990 to around 8% in 2005 (Figure 1). Other indicators of poverty also decreased considerably in this same interval. The intensity of poverty, measured by the poverty gap (FGT1), also declined during the period. Figures for this indicator fell from 3.3 to 2.3. The severity of poverty (FGT2) indicator also showed a significant reduction, falling from 1.45 to 0.96, indicating that the average distance to the poverty line for poor households decreased (World Bank 2013b). Across sub-regions, South Caucasus is the only sub-region where poverty rates increased considerably. In the rest of them poverty rates either fell, in some cases significantly, or increased only slightly. Figure 1. Changes in GDP per capita and Poverty Headcount in ECA (a) GDP (b) Poverty Headcount Source: World Bank (2013a) for panel (a) and World Bank (2013b) for panel (b). (*) ECA estimates are population- weighted. In spite of this positive economic performance, a large fraction of ECA's population – even among the best performers - reports their economic situation in the 2000s to be worse than in 1989. An analysis of data collected by the EBRD and the World Bank in 2006 (LiTS) on perceptions of economic well-being showed that the majority of households do not think they are better than before transition (EBRD 2006). When asked                                                              2  New EU Member States include Bulgaria, Croatia, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Poland,  Romania, Slovakia, and Slovenia.  3 | P a g e     about their current position at the time in the income distribution, most households located themselves in the lower steps (Figure 2). However, when asked about what position they recall being located in 1989, most households indicated a higher step. The most common movements from 1989 to 2006 are negative, and only around 15% of the households perceive an improvement. When asked again in 2010 about their relative position in 2006 and 2010, most households reported to perceive themselves in the lower steps, showing a distribution much similar to the one reported for 2006 before than for 1989. Figure 2. Ten-Step Distribution of Household Perceived Position in 1989 and 2006 (LiTS 2006) and 2006 and 2010 (LiTS 2010) for ECA (a) Step Levels 1989 and 2006 (LiTS 2006) (b) Step Changes 1989-2006 (LiTS 2006) (a) Step Levels 2006 and 2010 (LiTS 2010) (b) Step Changes 2006-2010 (LiTS 2010) Source: 2006 Life in Transition Survey (LiTS 2006) and 2010 Life in Transition Survey (LiTS 2010) (individual weights)3 Results are for the ECA region by pooling data across countries. The observed disconnect is striking if we consider that 2006 was a year of solid economic expansion in the region, close to the peak of the expansion that covered most of the 2000s. With commodity prices at a high level, and the world economy growing more than 3% on average in the last five years, 2006 observed one of the highest growth rates in the region. The GDP per capita average annual growth between 2000 and 2006 was 5.5%, and in 2006 all the countries in the region observed a positive growth. The average growth in 2006 across countries was more than 8%, with even 5 countries growing more than 10% (World Bank 2013a).                                                              3  The countries included are: Albania, Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina, Bulgaria, Croatia,  Czech Republic, Estonia, Georgia, Hungary, Kazakhstan, Kyrgyzstan, Latvia, Lithuania, Macedonia (FYR), Moldova,  Montenegro, Poland, Romania, Russian Federation, Serbia, Slovakia, Slovenia, Tajikistan, Turkey, Ukraine, and  Uzbekistan.  4 | P a g e     The reported low levels of perceived well-being are relevant in themselves, considering that subjective measures of well-being provide an important dimension to assessing individuals’ quality of life. The Commission on the Measurement of Economic Performance and Social Progress (Stiglitz, Sen and Fitoussi, 2009), for instance, emphasized that objective and subjective measures of well-being are relevant to assessing quality of life. Increasing subjective well-being is an end in itself, but it could influence people’s behavior, including support for reforms and labor market decisions. Therefore, understanding what drives people’s perceptions is extremely relevant. The 2006 World Development Report highlighted, for example, that income inequality influenced people’s happiness – regardless of absolute income levels- linked to perceptions of “unfair processes and unequal distribution of opportunities” in society (World Bank 2005, p.82). This study aims at documenting the patterns of perception of economic mobility in the region and across groups of countries, and to identify factors that are correlated with the different types of perceptions of economic mobility observed in the region. The analysis tests for the correlation of mobility with specific variables capturing fairness and trust in public institutions, inequality, personal and household characteristics, and macroeconomic performance indicators. Results confirm that factors beyond objective well-being are associated with the reported perceptions of mobility, as are the perception of trust and fairness in the country, the age and education of the respondent, and macro variables that capture the changes observed in the countries since the transition. Households in countries where income/consumption distribution has become more unequal have a more negative perception of their situation. Further, findings show that trust in public institutions, perceptions of fairness in the economy and preference for a market economy are associated with perceptions of improvement. Younger cohorts and more educated people are more likely to perceive improvements in their economic situation, likely linked to these groups’ ability to access the opportunities introduced by the reforms and today’s markets. An interesting but intuitive result is that a more severe reduction of the size of the State is also associated with more negative perceptions. This is likely capturing the intensity of the reforms applied in these countries and its consequences in terms of uncertainty and employment outcomes. Finally, macro variables also play some role. Countries with weak growth have on average more negative perceptions, and a more severe economic contraction during transition is associated with more negative perceptions. The rest of the paper is organized as follows. Section 2 presents the conceptual framework and key findings from the literature on which the analysis is based. Section 3 describes the data used in the analysis. Section 4 presents the main findings, looking at correlates of perceived economic mobility. It covers profiling, multivariate analysis of correlates with subjective mobility and disaggregation across sub-regions and low/middle income countries. Section 5 presents the conclusions of the study. 2. Conceptual Framework for the Analysis of Perceptions of Mobility Subjective measures of well-being, like “satisfaction with life” or Catril’s Ladders, complement the information about quality of life that objective measures provide. Economists have for long preferred to study revealed preferences, expressed by purchase decisions, rather than declarations of preferences. Under the typical microeconomic theory assumptions, revealed preferences are enough to infer individual’s well-being. However, a large amount of literature from behavioral economics and psychology find that people often make choices that do not necessarily increase their happiness. Their choices may hence not reflect their actual 5 | P a g e     preferences and focusing exclusively on actual choices may leave an incomplete picture to understand well- being. (Kahneman and Krueger 2006) There is solid evidence in the economic literature that the most common measures of subjective well- being are consistent enough in time to be a reliable measure of well-being, and are not mainly driven by mood changes or personal distortions. Krueger and Schkade (2008) analyzed the test–retest reliability of two types of measures of subjective well-being (a standard life satisfaction question and a measure based on recording of the satisfaction experienced at different times of the day) and obtained a serial correlation of about .60 when assessed two weeks apart. This is lower than the reliability ratios typically found for education, income and many other common microeconomic variables. Nonetheless, the estimated degree of reliability of subjective well-being data is high enough to detect effects when they are present in most applications, especially if samples are large and the data are aggregated across people or activities. (Krueger and Schkade 2008) Even though the focus of this paper is on perceptions of economic mobility, we borrow from the literature on perceptions of economic well-being to identify factors behind the mismatch between objective and subjective measures. For this, we followed the accounts of Selezneva (2011) and Dolan et al. (2008) about determinants of satisfaction with life (SWL). In her account, Selezneva (2011) reviews more than 70 studies that associate subjective well-being with income, work and family life, in the economic, psychological and political science literature, paying particular attention to transition countries. Dolan et al. (2008) reviews papers published in economics journals since 1990 as well as some key reviews in psychology that focus in finding individual-level, family-level and social-level determinants of subjective well-being. Both papers categorize the determinants roughly in income, personal characteristics, attitudes and beliefs, work conditions and macro environment. There is agreement in the literature that the correlation between income and subjective well-being is at most weak, leaving room for other factors to explain the perceptions of well-being. For instance, Dolan et al. (2008) notice that results in the literature generally suggest positive but diminishing returns to income. Zaidi et al. (2009) using the 2006 LiTS obtain a positive correlation between income and satisfaction, though Cojocaru and Diagne (2013) find only a weak correlation between objective welfare and the household relative income position using the the 2010 LiTS. Clark and Oswald (1994) in the UK, and Frey and Stutzer (2000) in Switzerland find at most a low correlation between income and subjective-well-being. Income inequality can affect negatively the perceptions of well-being when is associated with an unfair process and unequal distribution of opportunities. Using data for Europe and U.S. states, Alesina et al. (2004) obtains that even after controlling by income a higher national-level inequality is associated with more negative perceptions of well-being. However, it may also be the case that for countries with higher economic mobility inequality communicates opportunities, and hence higher inequality may signal a higher reward to effort, and have a positive effect on well-being (Dolan et al. 2008). The 2006 World Development Report highlights that income inequality influenced people’s happiness – regardless of absolute income levels- linked to perceptions of “unfair processes and unequal distribution of opportunities” in society (World Bank 2005, p.82). Finally, Abras (2012) obtain that circumstances determined by the lottery of birth (e.g. gender, parental status and politics), affect the perceptions of well-being when they play a role in determining income by means of facilitating positions or contracts. Perceptions of fairness are also correlated with higher levels of reported self-satisfaction. For instance, using data on Italian workers, Tortia (2008) obtains that worker well-being is crucially influenced by fairness concerns. He obtains that self-satisfaction is positively affected by how fair workers perceive are the procedures 6 | P a g e     within the organization (e.g. flow of information, promotions), even above how fair they perceive their compensation is with respect of their qualifications.. The author also obtains that public organizations are found to be at a disadvantage with respect to the private sector regarding both the degree of satisfaction and perceived fairness. In a related study, Krueger and Mas (2004) identify that workers perception of fairness can affect the quality of their work and the effort executed in performing their tasks. Finally, as World Bank (2005) notices, the experimental and subjective well-being literature in economics and social psychology points that there is something deep and fundamental about our taste for fairness and equity. Equity, the report indicates, matters intrinsically and fundamentally for human beings. Trust in other people and in national institutions plays an important role on the perceptions of individual well-being. Using individual-level data from different countries, Helliwell (2003) finds that trust in other people is associated with higher levels of well-being. Though it is possible that optimism produces higher assessments of both trust and well-being, he explains, if the answers relate to people with whom people interviewed interact, then the positive effect captures the benefits of living in a high-trust community. On another study using similar data, Bjørnskov (2007) found that generalized trust increases life satisfaction, and a social capital factor score (comprised from responses to generalized trust, civic participation and perceptions of corruption) are positively associated with life satisfaction. Finally, trust in key public institutions such as the police, the legal system and government is also found to be associated with higher life satisfaction (Helliwell & Putnam, 2004; Hudson, 2006). Personal characteristics (like age, gender and ethnicity) and socially developed characteristics (like education, health or employment status) also have an impact and affect the perceptions of well-being. For the first group, studies suggest a U-shaped curve with higher levels of well-being at the younger and older age points and the lowest life satisfaction occurring in middle age. There is also evidence that women and certain ethnic groups report higher levels of happiness (Dolan et al. 2008). For the second group, some studies find a positive relationship between each additional level of education and subjective well-being (SWB), while others find that middle-level education is related to the highest life satisfaction. Even so, it is difficult to isolate the effect of education due to unobservable variables (e.g. motivation, intelligence, family background) and the different channels through which education can affect happiness (e.g. health, income) (Dolan et al. 2008). Studies show a more consistent positive effect of physical and psychological health on happiness, and, inversely, a negative effect of unemployment on happiness (Dolan et al. 2008). Macroeconomic factors can also affect the perceptions of individual well-being when these factors have a direct impact on individuals’ life, as inflation or unemployment. National unemployment rates reduce subjective well-being (Alesina et al. 2004, Di Tella et al., 2001, 2003; Wolfers, 2003), most likely through the induced fear of personal unemployment, which arises from a higher national unemployment rate. Inflation also has a negative effect on well-being, as studies in Europe, Latin America, and the United States have found (Alesina et al., 2004; Di Tella et al., 2001, 2003; Wolfers, 2003, Graham & Pettinato, 2001), although the results are not conclusive, as many other possible correlates, like income inequality or lack of trust, are not controlled for in the regressions (Dolan et al. 2011). There is evidence that reforms implemented in former socialist countries have affected negatively the perceptions of well-being. Most authors agree in that populations of transition countries have paid for the socioeconomic transformation with their well-being. (Selezneva 2011) Reforms have introduced changes in all life dimensions, forcing the populations to face, among others, such new phenomena as unemployment and high fluctuation of incomes (Andren and Martinsson 2006). This affected the perceptions of well-being, which 7 | P a g e     in the initial years plummeted and saw the proportion of people in the least happy categories soar (Blanchflower and Freeman 1997; Lelkes, 2006a and 2006b ; Angelescu, 2008). By 2011, there is still a considerable gap in subjective well-being between transitional and no transitional countries, even though smaller than after the initial reforms (Lelkes, 2006b). 3. Individual and National Level Data The individual-level data used for this study comes from the Life in Transition Survey collected in 2006. This survey was collected by the European Bank of Reconstruction and Development (EBRD) and the World Bank (WB). Although the LiTs was conducted in 2010 as well, the study focuses on the 2006 round to capture perceptions in a period of economic boom – to avoid results being driven by the impact of the global economic crisis – and given the more comprehensive information on labor histories collected in the 2006 round. Perceptions on economic mobility in 2010 are, however, presented in annexes as well to complement the analysis. It collected information about socio-demographics characteristics and attitudes and values of the population in Eastern Europe, Central Asia and Turkey. It covered 29 countries, with a sample of 1,000 households per country.4 The respondent of the survey is a random member of the household interviewed older than 18 years old. The survey is divided in two sections. The first one was answered by the household head and contains the household roster and the variables on perceptions of economic mobility. The second part is answered by a randomly selected household member older than 18 years old, and contains information on employment, employment history since 1989 and an extensive set of questions on attitudes and values. We merged the information coming from the two respondents in the household. This implies we assume the answer given by the household head is representative of the household opinion. Personal and household characteristics used for the estimations are constructed based on the data available in the employment and household characteristics sections. We used this information to obtain the age, education, ethnicity, religion, and employment status of the respondent. Also, we used this section for the education of the father. We used the household characteristics to construct the location (urban/rural), household size, consumption and assets owned, and the shares of children and seniors. We also extracted information about the education of the father, as an indicator of condition that cannot be affected by the individual, to test whether they play a role in the perceptions of mobility. Attitudes and beliefs are obtained from the module of opinions included in the survey, which covered perceptions about the dynamics of the society and valuation of institutions. For the variable on trust in public institutions, we constructed an index based on trust in five public institutions (Presidency, Cabinet of Ministers, Parliament, Judges and Police). For each institution, the original variables were coded in 5 categories, ranging from distrust to full trust. We collapse this information in three categories and construct an index that can take three values (-1, 0, or 1) corresponding to distrust, neutral or trust in each institution. The aggregate index is the arithmetic average. For fairness, we used a question on what the respondent beliefs is rewarded in society, either effort and intelligence, or contacts and corruption/criminal ties. For market economy support, we took a question asking to express a preference between a market economy or a centrally planned one. Finally,                                                              4  It included Mongolia, EBRD country of operation since 2006, and part of the World Bank East Asia and Pacific  Region. Turkey is not an EBRD country of operation. Turkmenistan was not included.  8 | P a g e     for a perception of performance related to a local comparison group, we took a variable asking for the perceived performance compared to colleagues in 1989. Employment histories were constructed by taking advantage of the information collected on job market performance for each year between 1989 and 2006. The second section contains an extensive set of questions about life events and labor market history since 1989 to 2006, year by year. Unemployment may have been underreported, since working at least one part of the year is enough to consider the year as employed. If more than one position was held during one year, they are all reported in different cells. This information is used to construct the labor history of the interviewed person, including time worked in the public or private sector and number of different jobs held in this time. We restricted the information to the years when the interviewed person was between 18 and 65 years old.5 We also took advantage of the information on labor histories to construct a measure of public employment for every year between 1989 and 2006. Using the information on enterprise ownership, we were able to obtain the share of employment, excluding public administration, corresponding to the public sector. As can be expected, this share was considerably high in 1989 (with the exception of Turkey), and has fallen dramatically for most countries by 2006. The data at the national level was obtained from the World Development Indicators database (World Bank 2013). We used GDP per capita in real terms, Gini coefficients, and the ratio of the share of income the top 20% and the bottom 20%. In many cases the data for the years 1989 and 2006 was not available, so we replaced for the closest year available, with up to three years of difference. 4. Perceptions of Economic Mobility and Their Correlates a) What are the perceptions of economic mobility observed in ECA? A substantial share of households in the region perceives that their standards of living have fallen compared to 1989. In 2006, around 50% of households in the region declared that they saw their countries in worse economic standing than in 1989. Moreover, when asked about the specific situation of the household in the same period of time, again, around 50% of the households declared a worsening in their economic situation (Figure 3). The perceived fall in living standards is consistent across all sub-regions. Only households in Central Asian countries perceive in majority that their countries are doing better than in 1989 (Figure 3). This is indeed a region in which GDP per capita was almost 15 percent higher over the period. For all the other sub-regions – even the best performers in terms of GDP growth - the prevalent view is of a worsening of the national economic conditions, peaking at almost 70% in the Western Balkans. About the conditions of the household interviewed, the results are similar, and even more dramatic. On average, people across all regions consider that their household is doing worse than in 1989, ranging from 41% in New EU Member States (again, one of the best performing regions) to almost 60% in the Western Balkans.                                                              5  We considered Out of Labor Force the cases with no employment information between 1989 and 2006, or  decided not to work, or declared are officially retired & not worked & were at least 55 for women and 60 for men,  or declared left job because of retirement and never worked again or declared studying and not working. We  considered employed if worked for wages or self‐employed. Finally, we considered unemployed if received  unemployment benefits and did not working, or if it was part of the Labor Force, and did not have a job.  9 | P a g e     The majority of households located themselves in the lower steps of the income distribution in 2006. Household heads were asked where in a ten step ladder they would situate their household in 2006 (Cantril’s Ladder).6 In a perfect-knowledge world, this should result in a uniform distribution, with households identifying their correct decile immediately. However, this is not the case. The distribution seems to be capturing the perceived distance to the highest income household, hence resulting in a distribution similar to a log-normal one. The majority of households in ECA report themselves to be in the steps 5 and below (Figure 4, panel a). This results persisted when a different sample of households were asked the same questions in 2010, as part of the second round of the LiTS. Most households then recalled being in the lower steps of the distribution in 2006, and still being around those steps in 2010. Accordingly, three quarters of the households reported no movements across steps or at most a change in one step between 2006 and 2010 (Figure A1.4). Figure 3. Satisfaction with Economy and Household Living Standards (2006 vs. 1989) (a) Economic situation in country better in 2006 than 1989 (b) My household lives better in 2006 than in 1989 Source: 2006 LiTS (individual weights) However, most of the households also recall being in the middle steps of the income distribution in 1989. When households were asked about their position in a ten-step income ladder in 19897, the answers this time concentrated around the center steps (Figure 4, panel b). Around 50% of the respondents locate                                                              6  The literal question asked was: “Please imagine a ten‐step ladder where on bottom, the first step, stand the  poorest and on the highest step, the tenth, stand the richest. On which step is your household today?”  7  The question asked was: “Now imagine the same ten‐step ladder around 1989, on which step was your  household at that time?”  10 | P a g e     themselves in the steps 4, 5 and 6, and the distribution of the answers shows a symmetric shape, in sharp contract with the right-skewed distribution from 2006. These distributions reflect well what income distributions can be expected from a market and a central-planned economy. The right-skewed distribution obtained from the 2006 answers is consistent with income distributions observed in market economies. The symmetric distribution obtained for 1989 is consistent with centrally-planned economies where equality among its citizens is highly valued. Consequently, close to half of the households in the region perceived a fall in their relative position in the Cantril’s ladder. Pairing the answers on the ladder steps in 1989 and 2006 can provide information on perceptions of mobility in this period. Using as a definition of movement of a change in at least two steps in any direction, 47% of the households in the region declared to have moved down in the income distribution, regardless of their starting step (Figure 5). Around 40% of households perceive they have stayed in the same step, or moved at most one step in any direction. Figure 4. Household Perceived Position in 10-Step Income Ladder (2006 and 1989) (a) 2006 (b) 1989 Source: 2006 LiTS (individual weights) Less than 13% of households in the region perceive an improvement in their relative position. Close to 90% of the households who perceived an improvement are households who were located in the 5 lowers steps of the income distribution in 1989, and moved up to steps 4 to 8. Only 7% of the households who perceived an improvement moved all the way up to the steps 9 or 10 in 2006. This pattern of perceived downwards mobility is common across all sub-regions (Figure 5). With the exception of New EU Member States and Turkey – the ones with higher economic growth during the period - in all sub-regions the most prevalent perception is of downwards mobility. And even in these countries, almost 40% of the households declared to have moved down, behind around 50% of the households that consider doing only as well as in 1989. Western Balkans and South Caucasus are the regions with the most negative perceptions, where households that feel they have moved down reach 60% and 65% respectively. In the Russian Federation, Moldova, Ukraine and Belarus, and Central Asia countries, this share of households is close to 50% as well. 11 | P a g e     Figure 5. Perception of Economic Mobility 1989 – 2006 (a) ECA (b) New EU Member States (c) Western Balkans and Turkey ECA: Type of Inter-step Mobility 1989 to 2006 EU10+1 & Turkey: Type of Inter-step Mobility 1989 to 2006 Western Balkans: Type of Inter-step Mobility 1989 to 2006 70 70 70 59.9 60 60 60 50 50 50 46.5 47.5 Percentage Percentage Percentage 40.6 38.7 40 40 40 29.2 30 30 30 20 20 20 12.8 13.8 10.9 10 10 10 Down Same Up Down Same Up Down Same Up (d) Russia, Moldova, Ukraine (e) South Caucasus (f) Central Asia and Belarus CIS: Type of Inter-step Mobility 1989 to 2006 South Caucasus: Type of Inter-step Mobility 1989 to 2006 Central Asia: Type of Inter-step Mobility 1989 to 2006 70 70 70 65.0 60 60 60 50.5 50 50 50 47.3 41.5 Percentage Percentage Percentage 40 40 40 36.7 30 30 30 26.4 20 20 20 12.8 11.2 8.6 10 10 10 Down Same Up Down Same Up Down Same Up Source: 2006 LiTS (individual weights) b) What is the socio-demographic profile of those with perceptions of downward mobility? Perceptions of mobility vary by socio-demographic characteristics of the households, though with very few exceptions the prevalent perception is to have moved down in the income distribution after the transition. We performed a descriptive analysis of the perceptions by income, age, location, education, and employment status. This analysis is aimed at shedding light into the postures of different socio-demographic groups, rather than to find causality or to isolate the effect of particular characteristics. The results show differences in the intensity of the perception, although in almost all cases the predominant view is to have moved down. Also, we would ideally use the characteristics before and after transition for this profiling, but the only information available about characteristics is from 2006. We prefer to report a profile based in 2006 information since this is still informative. Not surprisingly, poorer households in 2006 have more negative perceptions of mobility. Using two different measures of welfare at the household level, we consistently obtain that poorest households in 2006 are more likely to report having moved down in the income distribution (Figure 6). The measures used are monthly household consumption and an index based on asset ownership. Results are roughly similar for both measures. The share of households that perceive movements down are around 55% for the lowest quintile and 12 | P a g e     less than 40% for the richest quintile. In contrast, the share of households perceiving movements up is around 7% for the poorest quintile and around 20% for the richest. There is not much variation across quintiles in the shares of households reporting staying in the same step. Older people report the worst perception of economic mobility. Figure 7 (panel a) shows that the share of young respondents (18-34) reporting movements down is only 35%, while this rate reaches 60% for the oldest group (66+). For the movement up category we observe the opposite pattern, while more than 17% of the young respondents perceive movements up, only 6% of the oldest respondents think the same. There is also considerable variation for “staying in the same step” category: almost 50% of the youngest against 35% of the oldest group. In many cases, the respondent to the employment section, from where the age information is collected, is different than the respondent to the perceptions section. However, results when restricting the sample to cases where the same person responded are very close to the one here reported, and even the differences are slightly more accentuated. These results are available in the Appendix A4. Figure 6. Perception of Economic Mobility 1989 – 2006 by Welfare Level (a) Based on Consumption (b) Based on Asset Index Source: 2006 LiTS (individual weights) Perceptions do not vary considerably across areas of residence. Dividing the sample in three groups according to the degree of urbanization (metropolitan, urban and rural) does not result in marked differences in the responses to perception (Figure 7, panel b). The share of households perceiving a movement down vary from 43% for the metropolitan areas to 48% for the rural, and for the movement up goes from 16% for the metropolitan to 10% in the rural.8 Low-educated people report the most negative outlook regarding upward mobility, while tertiary education are the less likely to report downward mobility. Figure 7 (panel c) shows that for people with no education or only primary education, the share of households reporting movements down reaches 47%, and for the secondary, or professional education this share is 48% and 50%. However, for college-educated this share goes down to 40%. For the movements up, the differences in shares is slightly more pronounced among levels of education: 10% for no or primary education, and 13%, 12% and 18% for the people with secondary,                                                              8  The labeling of the category can be misleading, since the three categories were defined for all countries and even  countries with small populations and no big cities have considerable shares of population in the metropolitan  category. However, we decided to keep this categorization to respect the original categorization of the survey, and  because regardless of the definition of each category by country, there are no considerable variations in the  answers.  13 | P a g e     professional or university education, respectively. Restricting the sample to cases where the respondent to education and perceptions are the same show similar results (Appendix A4). People employed at the moment of the survey have the less negative perception of mobility compared to the jobless, though the predominant perception is still negative. The share of respondents that perceive movements down in the income distribution is 42% among the employed, while this share is more than 50% for the unemployed and inactive (Figure 7, panel d). This difference is also clear in the movements up across the income distribution. While for the employed this share is 16%, for the unemployed and inactive this share is less than 10%. Again in this case, when restricting the sample to the cases where the same person responded employment and perceptions, results are qualitatively equivalent. (See Appendix A4.) Figure 7. Perception of Economic Mobility 1989 – 2006 by Socio-demographic Characteristics (a) By Age (b) By Location (c) By Level of Education (d) By Employment Status Source: 2006 LiTS (individual weights) c) What variables are correlated with perceptions? A multivariate analysis In this study we carried out a multivariate analysis to identify the joint effect of individual variables on perceptions of economic mobility, with a particular focus on variables on fairness and trust. The ideal setting for empirically testing causality in the social sciences is when there is strict exogenous variation in one variable that can be associated to the dependent variable. Unfortunately, we cannot make such a strong claim in our data. What we do, in contrast, is testing the statistical association of a set of variables with the perceptions of economic mobility. The results are presented below. 14 | P a g e     Regression results are presented in the Appendix and are centered on identifying correlates of downward mobility between 1989 and 2006. The dependent variable takes a value of one if the individual reported moving up or staying in the same position, and zero if it reported downward mobility. We estimated probit regressions to facilitate the interpretation of the results; however, results were qualitatively equivalent under a multivariate regression on three types of movements. A movement is defined (as in section 3) as a movement in any direction of at least two steps, regardless of the initial point. Movements of only one step are considered as staying in the same step. We excluded households in the two lowest and uppermost steps in 2006 from the sample to avoid biases introduced by the mechanical effect of being unable to move in all possible directions (move down for the upper steps in 2006, and move up for lowest steps in 2006). All regressions and a thorough discussion of the methodology used are presented in the Appendices. Perceptions of inequality, fairness and trust matter Increases in country income inequality are associated with more negative perceptions of economic mobility. Table A2.2 shows that household in countries in which inequality increased between 1989 and 2006 – measured by the ratio of the 80th/20th percentiles– had a higher probability of reporting downward movements in the income distribution (Column (3)). Using an alternative measure of Gini, obtained from the World Income Inequality Database shows also a negative effect of the increases in inequality with respect to the perception of movements up9 (Column (4)). This alternative measure, though, is less consistent than the obtained from the World Development Indicators and is incomplete for most former Yugoslavia countries, so we preferred to use the original from the WDI for the rest of the report. As reported earlier when citing the 2006 WDR, “One reason that inequality might make people less happy, even when controlling for absolute income levels, is that it violates their sense of fairness” (p. 82). The Gini, a less sensitive measure to changes in the tails of the distribution, can potentially not capture this correlation. Perceptions of fairness in moving ahead in life are associated with more positive perceptions about economic mobility. The perception that effort, hard work and intelligence are the most important factors to succeeding in life, rather than criminal or corrupt ties, or political connections, reduces the probability of reporting movements down. Specifically, those who believe in fairness have 9 percentage points less probabilities of perceiving a worsening in their relative position. This is particularly the case in higher income countries in ECA (Table A2.3). Trust in public institutions is associated with more positive perceptions about economic mobility. For any individual, moving from a neutral approach to trust to becoming trusting in public institutions reduces the likelihood of perceiving movements down by 7 percentage points. This effect is stronger (10 percentage points) in low and lower-middle income ECA countries. 10 The perception of having done better than colleagues in 1989 is associated with less probabilities of perceiving downward economic mobility. Individuals who reported a relative improvement compared to their peers pre-transition have a 4 percentage point lower probability of reporting downward mobility. This                                                              9  http://www.wider.unu.edu/research/Database/en_GB/database/   10  We also performed a robustness check, by replacing this indicator for trust in religious institutions. The rationale  behind is that if people who are more likely to be trustful are most likely to perceive improvements in their  situation, then trust in religious institutions should be significant too. The results show that this variable is not  significant. This suggests that the connection between trust in public institutions and perceptions is not driven by  individual biases, but by actual beliefs about these institutions.  15 | P a g e     effect is stronger and significant for low and lower-middle income ECA countries (10 percentage points), but not significant for countries with higher incomes in the region. Robustness checks, however, raise concerns about this finding. 11 Support to market economy is also positively correlated with more positive perceptions of economic mobility. Declaring support to a market economy, as opposed to a centrally-planned one, is associated with 7 percentage points fewer chances of perceiving a worsening in the position in the income ladder. This effect may be partially driven by being precisely the ones who have benefited from the new conditions introduced in a market economy (for instance, rewards to entrepreneurship, risk-taking behavior, or high tolerance to uncertainty), the ones who perceive improvements in their relative well-being. Robustness checks suggest that actual belief in the market system, rather than a general attitude or bias towards changes introduced after transition, is what is driving the association.12 The young and more educated fare better Young populations are more likely to declare improvements in their economic position. Consistently with the results in the profiling section, age is strongly associated with the probabilities of perceiving improvements in well-being. Belonging to the 18-35 years old category, as opposed to any of the other age groups decreases the probabilities of perceiving a worsening in the economic situation of 7 percentage points. These results contradict the common finding in the literature of a more optimistic perception of life for older people (Dolan et al. 2008). Given that the youngest cohort (18-35 in 2006) was under 18 years old in 1989, the result for this group may be capturing also intergenerational perceived mobility. We performed some robustness checks by restricting the sample to cases with the same respondent in the employment and perception sections, and to the cases with only workers older than 35 years old (18 in 1989). Results are presented in the Appendix (Table A3.2) and show results largely consistent with the original full sample, although the age now has a positive effect on perceptions of economic mobility.13 More educated people are more likely to declare improvements. College-educated respondents are less likely to perceive their position falling in the income distribution, as opposed to the other levels of education.                                                              11  We restricted the sample only to the respondents who think their household is doing worse than in 1989, in  absolute terms. Among these respondents, perceiving that they are doing better than the 1989 colleagues would  be most likely driven by actual perceptions, rather than by overall optimistic views about life. The same happens  when the sample is restricted the cases that think they are doing better or at least the same than in 1989, in  absolute terms. In neither of both cases the perception of doing better than the comparison group is significant,  raising concerns about the driver behind the initially reported correlation. Most likely, the original result is driven  by common idiosyncrasies contained in the perceptions of mobility and relative performance, rather than actual  perceptions of improvements.    12  We also checked the robustness of this result using the opinion about democratic against an authoritarian  regime. If the support to a market economy is driven by support in general to the changes introduced with the  reforms, most likely the support to a democratic regime would be also correlated with the perceptions of mobility.  However, support to democracy is not significantly correlated  13  The only noticeable difference is in the significance of belonging to the 55‐64 age group. The reported negative  effect over the probabilities of perceiving downward mobility turns statistically significant, though only weakly.  Also, the share of time spent working in the public sector becomes significant, showing now that more time in the  public sector between 1989 and 2006 is associated with more probabilities of perceiving a movement down. All  the other parameters sign and significance remain unaltered.  16 | P a g e     Moving to college-level education from any other category lowers the probabilities of perceiving falls by 6 percentage points. In today’s market economy and process of ongoing reforms, the young and more educated have a better perceived outlook. The perceived improvement in economic situation of these groups is likely linked to their relatively higher readiness – in terms of skills, for example – to benefit from the opportunities introduced by the reforms. This finding is consistent with the discussion in Arias et al. (2014) in which is pointed that across the region, an increasing share of employers cited skills as a major constraint to firms’ growth. In a context where most education and training systems have failed to keep up with the fast-changing labor market, being flexible and ready to acquire and develop non-routine higher-order skills, including socioemotional (“soft”) skills, translates into more job opportunities and higher earnings. Family background plays a role in perceptions The education of the father plays a role in determining the perception of mobility, hinting that opportunities in the country are associated with factors beyond the individual’s control. Including the education of the father in the regression shows that secondary and professional education lead to better perceptions of economic mobility than basic or no education. Surprisingly, a father with higher education does not lead to better perceptions than a father with primary education. These results signal that factors beyond the control of the worker may be driving the access to opportunities or the satisfaction in life. In case perceptions were to depend only factors that the interviewee can decide, father characteristics should not have any effect. Other variables, like father belonging to the communist party, did not show statistically significant results. Individual labor market experiences have a role, but less than expected Working for a public company and being self-employed are associated with lower perceived downward mobility. Public workers are more likely to report improvements in the economic situation, though this effect is statistically weak. This result certainly contradicts previous findings in the literature (e.g. Tortia (2008)), although previous results have focused in developed countries and did not analyzed periods with dramatic changes comparable to the transition lived in the ECA region. Considering the generalized reduction in the size of the state across countries in the region and the strong preference for public jobs14 (EBRD 2010), being currently employed for the public sector may indicate that these workers were able to retain these positions and associated benefits. Hence, it is intuitive for them to perceive an improvement in their situation since, for many of their peers, reforms involved considerable undesired changes. In the case of self-employment, this can be signaling entrepreneurial ability and willingness to take risks, values highly regarded in a market economy. Surprisingly, labor history does not play any significant role on the perceptions of economic mobility. The share of time in the labor force spent working for the public sector is negatively correlated with upward mobility, although the degree of statistical significance is weak. The share of time working for the private sector does not have a significant effect over the perceptions. When the sample is restricted to household heads and older than 35 years old workers the share of time employed in the public sector shows a positive effect over the probabilities of perceiving downward mobility (Table A3.2). The variable is defined as shares of time spent                                                              14  According to data collected in the LiTS 2010 (EBRD 2010), 45% of workers 18+ in the region would prefer to work  for a state‐owned enterprise (30%) or the public administration (15%). Being self‐employed comes behind with  25% of the sample, 20% would prefer in a private firm, and 10% does not know. While these results hide some  regional disparities, it is clear also that it highlights a preference for the types of positions offered in the public  sector.   17 | P a g e     either on public sector, private sector or unemployed/out of labor force (base category), hence the effect is with respect to being unemployed/inactive. This is a somehow puzzling result, since there is weak evidence that being employed by the public sector at the time of the survey has the opposite effect (less chances of perceiving downwards movements). A possible explanation is that this variable is picking up the effect of belonging to the 55-65 years old (in 2006) group. (See Figure A1.2) It is in this group where the shares of time in public sector reaches its highest, as opposed to the younger workers (private) and older workers (unemployed/inactive). For the number of jobs held between 1989 and 2006, which can work as a metric of the turmoil introduced by the reforms in economies where the labor market used to guarantee a position for all workers for life, the effects are also non statistically significant. This result hints that the current personal employment situation is more important than the memory for the perceptions of mobility. The role of income and macroeconomic indicators Current income explains only a minor fraction of the observed variation in perceptions. Estimating OLS regressions on the raw changes in step between 1989 and 2006 against two welfare measures (consumption and asset index), we obtain that welfare explains only a minor fraction of the variation in perceptions, though the parameters are statistically significant. Consumption can only explain 5% of the variations in perceptions, while the asset index explains only 8%, as measured by using the R-squared of the regressions (Table A2.1). Moreover, when consumption deciles are included in the regression containing all the other variables, they do not show any significant effect over perceived economic mobility. Countries with weak growth have on average more negative perceptions, particularly for those countries faring better pre-transition. A 10-percent increase in GDP per capita between 1990 and 2006 translates into 3 percentage points higher probabilities of perceiving a movement up. However, the starting point matters for perceptions of mobility: for countries with similar growth, a higher GDP per capita in 1990 (by US$1,000) translates into a 3 percentage point lower probability of reporting improvements. A stronger economic contraction during the transition period is associated with more positive perceptions. After controlling for the effect of initial GDP per capita level and the change between 1990 and 2006, a more severe contraction of the GDP per capita during the nineties is associated with more probabilities of self-reported upward economic mobility. This result suggests that those in countries that experienced tough economic times in the nineties place more value in the country’s or own relatively improved economic situation by 2006. The lowest point during the transition period is likely to be a reference point for countries that had severe economic contractions. Bigger public sectors in 1989 are associated with more negative perceptions. With the exception of Turkey, in all the countries of the region the State controlled considerable productive resources in the economy. This translated in massive public employment which ranged from 57% for Czech Republic to 91% in Belarus (own estimation based on EBRD 2006). The regression analysis shows that an increment of 10 percentage points in the share of public employment in 1989 translates in 9 percentage points less probabilities of perceiving upward mobility. This result hints that populations in countries that use to have more certainty about future employment may have been hit worse by the changes introduced with the liberalization of the economy. Accordingly, countries that reduced more intensely their share of public employment have more negative perceptions. Between 1989 and 2006 all countries in the region reduced the participation of the State in the economy. This ranged between 13 percentage points in Belarus to 50 percentage points in Bulgaria (own estimation based on EBRD 2006). A reduction of 10 percentage points in this share lowers the 18 | P a g e     probabilities of perceive a movement up in 3 percentage points. If we consider the reduction in public employment as a proxy of the intensity of the reforms applied in the economy, this result signals that, keeping all the rest constant, more intense reforms had a negative effect on the perception of well-being. This is an intuitive result given the uncertainty introduced in the lives of citizens during the transition. This effect, though, could be counterbalanced as reforms pay off in terms of growth and employment. Figure 8. Aggregated Effect of Country-level Variables on Perceptions of Economic Mobility 1989 – 2006: Probabilities of Downward Movements Source: Own calculations. Variables considered: GDP, Gini coefficient and share of public employment, 1990 and 2006 for all of them, plus the lowest GDP per capita between 1990 and 2006 as a percentage of the 1990 GDP per capita. All other variables considered as sample means. The combined effect of country-level variables on perceptions of downward economic mobility is below the regional average for the New EU Member States and Turkey, while for South Caucasus, Western Balkans and Central Asia is above the average (Figure 8). For this estimation we predicted the probabilities at the country level, using the aggregate variables in levels, and all the other variables at the regional sample mean. This way we tried to isolate the effect of the aggregate variables, while still obtaining economically significant results. The variables considered were GDP, Gini and public employment in 1990 and changes to 2006. The average probabilities of perceiving a movement down under this setting is 0.41 for the region. However, this result hides considerable heterogeneity. New EU Member States benefited clearly from their performance, as their predicted probabilities are in almost all cases below the regional average. Georgia and Armenia in the South Caucasus show a strong negative effect, while Serbia and Macedonia (FYR) in the Western Balkans have also considerable negative effects. All the Central Asia countries included in the analysis have also a negative effect, being all of them located above the regional average. 19 | P a g e     d) Are perceptions similar across sub-regions? The magnitude and significance of the results vary across sub-regions. This section explores the regression analysis in five country groupings in ECA: New EU Member States and Turkey; Western Balkans; Belarus, Moldova and Ukraine; South Caucasus Countries, Central Asia and Russia. Table A2.4 presents the results. Overall, the sub-regional results are consistent with those of ECA, though some of the most striking exceptions arise in the Western Balkans. Across most sub-regions, households who believe in fairness and that trust in public institutions have a less grim perception of their economic mobility. The exception is Central Asia, where the results are not significant. The effect of believing in effort, hard work and intelligence for succeeding in life is particularly strong in Russia, leading to a decrease in the likelihood of reporting downward mobility by around 16 percent. Trust in the presidency, cabinet of ministers, parliament, judges and the police decreases perceptions of downward movement across sub-regions. The result is not significant for Russia but it is strong and significant for all others, particularly for New EU Member States and Turkey. For the majority of advanced reforms and late reformers in the region, in contrast to intermediate reforms, support to market economy is positively correlated with higher subjective economic mobility. These are countries in which the reform effort started earlier and had already started to pay off (advanced), or in which the reform efforts are lagging behind and significant and necessary policy changes remain to be done (this reflection borrows from the typology in Arias et al, 2013). In intermediate reformers, the majority of which are from the Western Balkans and South Caucasus, support to market economy has no impact on mobility perceptions. In Central Asia and the South Caucasus, the perceived position relative to pre-transition colleagues matters. In these two sub-regions, the perceptions of being better off than colleagues in 1989 has a very strong association (around 13 percentage points) with moving up or at least staying in the same position. The Western Balkans show unique results in demographics and education. First, even if the older populations across sub-regions more often perceive downward movements, the Western Balkans have the strongest association between old age and negative perceptions of economic mobility: the probabilities of perceiving movements down are 10 percentage points higher for those ages 54-65 years and 14 percentage points higher for those over age 66 years. Second, in this same sub-region, advanced education, in contrast with the aggregate results, is associated with more negative perceptions on economic mobility. In other words, the more educated have a more pessimistic outlook on quality of life than those with little or no formal education. This education effect is also observed in the South Caucasus countries. This is not surprising if we consider the high unemployment rates for these regions for workers with higher education. For instance, in Macedonia, FYR, the unemployment rate for higher education graduates in 2006 was close to 20%, a similar rate than the one observed in Armenia and Georgia around those years15 (ILO 2013). Labor market status’ association with perceived mobility shows mixed results across sub-regions. The Western Balkans again stands out: the unemployed in this sub-region – one with particularly high unemployment rates- have more negative perceptions of economic mobility (at 8 percentage points); this effect                                                              15  Unemployment rate for Armenia in 2007 for persons with tertiary education was 20.4, while in Georgia it was  19.3 for the same year. (ILO 2013)  20 | P a g e     is, surprisingly and uniquely, also observed among those employed in the public sector in these countries16. Conversely, in New EU Member States and Turkey, in particular, public workers have a less negative perception of their economic mobility. Central Asia is the only sub-region in which private employment is associated with a more positive perception on mobility, and in which the opposite happens for working in the agricultural sector17. 5. Conclusions A substantial share of households in the region perceives that their standards of living in 2006 have fallen compared to 1989. Using as a definition of movement a change in at least two steps in any direction, 47% of the households in the region declared to have moved down in the income distribution, regardless of their starting step. Around 40% of households perceive they have stayed in the same step, or moved at most one step in any direction. Less than 13% of households in the region perceive an improvement in their relative position. This pattern is consistent across sub-regions. With the exception of New EU Member States and Turkey – the ones with higher economic growth during the period - in all sub-regions the most prevalent perception is of downward mobility. And even in these countries, almost 40% of the households declared to have moved down, behind around 50% of the households that consider doing only as well as in 1989. Western Balkans and South Caucasus are the regions with the most negative perceptions, where households that feel they have moved down reach 60% and 65% respectively. In Russia, Moldova, Ukraine and Belarus, and Central Asia countries, this share of households is close to 50% as well. Less affluent, older, rural and unemployed people show a more negative perception of economic mobility. Taking advantage of the information available in the survey for 2006, we obtained that these characteristics were associated with perceptions of downward economic mobility. Current income can only explain only a small fraction of the observed variation in perceptions. Welfare explains only a minor fraction of the variation in perceptions. Consumption can only explain 5% of the variations in perceptions, while the asset index explains only 8%, as measured by using the R-squared of the regressions Factors beyond objective well-being, including those related to trust and fairness are associated with perceptions of mobility. Households in countries where income distribution has become more unequal have a more negative perception of their situation. Further, findings show that trust in public institutions, perceptions of fairness in the economy and preference for a market economy are associated with perceptions of improvement.                                                              16  In the Western Balkans, the public sector employs 48% of the workforce, close to the 47% regional average.  17  In Central Asia, private employment accounts for 46% of the employment, below the regional average of 53%.  The agricultural sector employs a share considerably higher than the regional average: 31% of employed workers  vs. a regional average of 10%.  21 | P a g e     Younger cohorts and more educated people are more likely to perceive improvements in their economic situation. This result is likely linked to these groups’ ability to access the opportunities introduced by the reforms and today’s markets. A more severe reduction of the size of the State is also associated with more negative perceptions. This is likely capturing the intensity of the reforms applied in these countries and its consequences in terms of uncertainty and employment outcomes. National level variables are significantly associated with the perceptions of economic mobility. The combined effect of macroeconomic variables on perceptions of downward economic mobility is below the regional average for the New EU Member States and Turkey, while for South Caucasus, Western Balkans and Central Asia it is above the average. The magnitude and significance of the results vary across sub-regions. Across most sub-regions, households who believe in fairness and that trust in public institutions have a less grim perception of their economic mobility. For the majority of advanced reforms and late reformers in the region, in contrast to intermediate reforms, support to market economy is positively correlated with higher subjective economic mobility. In Central Asia and the South Caucasus, the perceived position relative to pre-transition colleagues matters. The Western Balkans show unique results in demographics and education. First, even if the older populations across sub-regions more often perceive downward movements, the Western Balkans has the strongest association between old age and negative perceptions of economic mobility. Second, in this same sub- region, advanced education, in contrast with the aggregate results, is associated with more negative perceptions on economic mobility. Labor market status’ association with perceived mobility shows mixed results across sub-regions. The Western Balkans again stands out: the unemployed in this sub-region – one with particularly high unemployment rates- have more negative perceptions of economic mobility. Conversely, in New EU Member States and Turkey, in particular, public workers have a less negative perception of their economic mobility. Central Asia is the only sub-region in which private employment is associated with a more positive perception on mobility, and in which the opposite happens for working in the agricultural sector. 22 | P a g e     References Abras, Ana; Hoyos, Alejandro; Narayan, Ambar and Sailesh Tiwari (2012) Inequality of Opportunities in the Labor Market: Evidence from Life in Transition Surveys in Europe and Central Asia. Manuscript. Alesina, A., Di Tella, R., & MacCulloch, R. (2004). Inequality and happiness: Are Europeans and Americans different? Journal of Public Economics, 88, 2009–2042. Andren, D., Martinsson, P., 2006. What contributes to life satisfaction in transitional Romania? Review of Development Economics 10, 59–70. Angelescu, L., 2008. The effects of transition on life satisfaction in Poland. In: Working Paper G08, Center for the Study of Democracy, Irvine, CA. Arias, Omar S., Carolina Sánchez-Páramo, María E. Dávalos, Indhira Santos, Erwin R. Tiongson, Carola Grun, Natasha de Andrade Falcão, Gady Saiovici, and Cesar A. Cancho (2014). Back to Work: Growing with Jobs in Europe and Central Asia. Washington, DC:World Bank. Bidani, Benu, Mame F. Diagne and Salma Zaidi (2012) Subjective Perceptions of the Global Economic Crisis in Europe and Central Asia. Policy Research Working Paper 5526. The World Bank, Washington, DC. Bjørnskov, C. (2007). The multiple facets of social capital. European Journal of Political Economy, 22, 22–40. Blanchflower, D, Freeman, R., 1997. The attitudinal legacy of communist labor relations. Industrial and Labor Relations Review 50, 438–459. Bogomolova, Tatyana and Vera Topilina (1999) Income Mobility in Russia in the mid-1990s. Economic Education and Research Consortium WP No. 99/11. Clark, Andrew E. and Andrew J. Oswald (1994) Unhappiness and unemployment. The Economic Journal (104, Number 424): 648–659. Clark, Andrew E.; Frijters, Paul; and Michael A. Shields (2008) Relative Income, Happiness, and Utility: An Explanation for the Easterlin Paradox and Other Puzzles. Journal of Economic Literature (46):95–144. Cojocaru, Alexandre (2011) Inequality and Well-Being in Transition Economies: A Non-Experimental Test of Inequality Aversion. Society for the Study of Economic Inequality Working Paper Series, WP 2011–238. Cojocaru, Alexandre (2012) Prospects of Upward Mobility, Inequality of Opportunity and Preferences for Redistribution in Transition Economies. Manuscript. Cojocaru, Alexandre and Mame F. Diagne (2013) How reliable and consistent are subjective measures of welfare in Europe and Central Asia? Manuscript. Dabalen, Andrew and Saumik Paul (2011) History of Events and Life-satisfaction in Transition Countries. Policy Research Working Paper 5526. The World Bank, Washington, DC. Denisova, Irina; Markus Eller; Timothy Frye and Ekaterina Zhuravskaya (2009) Who Wants To Revise Privatization? The Complementarity of Market Skills and Institutions. American Political Science Review (103, No. 2): 284–304. 23 | P a g e     Denisova, Irina; Markus Eller; Timothy Frye and Ekaterina Zhuravskaya (2011) Everyone Hates Privatization, but Why? Survey Evidence from 28 Post-Communist Countries. Manuscript. Di Tella, R., MacCulloch, R., & Oswald, A. (2001). Preferences over inflation and unemployment. Evidence from surveys of happiness. The American Economic Review, 91(1), 335–341. Di Tella, R., MacCulloch, R., & Oswald, A. (2003). The macroeconomics of happiness. The Review of Economics and Statistics, 85(4), 809–827. Dolan, Paul; Peasgood, Tessa and Mathew White (2008) Do We Really Know What Makes Us Happy? A Review of the Economic Literature on the Factors Associated with Subjective Well-Being. Journal of Economic Psychology (29): 94-122. Easterlin, Richard A. (1974) Does Economic Growth Improve the Human Lot? In: Paul A. David, Melvin W. Reder (Eds.), Nations and Households in Economic Growth: Essays in Honour of Moses Abramovitz. Academic Press, Inc., New York. Easterlin, Richard A. (2009) Lost in Transition: Life Satisfaction on the Road to Capitalism. Journal of Economic Behavior & Organization (71): 130-145. EBRD (European Bank for Reconstruction and Development). 2006. Life in Transition Survey. [Data file] Available from: http://www.ebrd.com/pages/research/publications/special/lifeintransition.shtml EBRD (European Bank for Reconstruction and Development). 2010. Life in Transition Survey. [Data file] Available from: http://www.ebrd.com/pages/research/publications/special/transitionII.shtml Ferrer-i-Carbonell, Ada (2005) Income and Well-Being: An Empirical Analysis of the Comparison Income Effect. Journal of Public Economics (89): 997–1019. Frey, Bruno S. and Alois Stutzer (2000) Happiness, economy and institutions. Economic Journal (110): 918– 938. Frey, Bruno S. and Alois Stutzer (2001) Happiness and Economics : How the Economy and Institutions Affect Human Well-Being. Princeton: Princeton University Press. Frey, Bruno S. and Alois Stutzer (2002) What Can Economists Learn from Happiness Research? Journal of Economic Literature (40, No. 2): 402–435 Graham, C., & Pettinato, S. (2001). Happiness, markets and democracy: Latin America in comparative perspective. Journal of Happiness Studies, 2, 237–268. Graham, Carol and Stefano Pettinato (2002) Frustrated Achievers: Winners, Losers and Subjective Well-Being in New Market Economies. The Journal of Development Studies (38, Issue 4): 100–140. Gruen, Carola and Stephan Klasen (2012) Has Transition Improved Well-being? Economic Systems (36): 11– 30. Hamermesh, Daniel (2004) Subjective Outcomes in Economics. NBER Working Paper Series, WP 10361. Cambridge, MA. 24 | P a g e     Helliwell, J. F. (2003). How’s life? Combining individual and national variables to explain subjective well-being. Economic Modelling, 20, 331–360. Helliwell, J. F., & Putnam, R. (2004). The social context of well-being. Philosophical Transactions of the Royal Society London, 359, 1435–1446. Hudson, J. (2006). Institutional trust and subjective well-being across the EU. Kyklos, 59(1), 43–62. ILO (International Labour Organization). 2013. Key Indicators of the Labor Market (KILM). 7th ed. Geneva, Switzerland: International Labour Office. Kahneman, Daniel and Alan B. Krueger (2006) Developments in the Measurement of Subjective Well-Being. Journal of Economic Perspectives (20, Number 1): 3–24. Krueger, Alan B., and Alexandre Mas. 2004. “Strikes, Scabs, and Tread Separations: Labor Strife and the Production of Defective Bridgestone/Firestone Tires.” Journal of Political Economy 112(2):253–89. Krueger, Alan B. and David A. Schkade (2008) The Reliability of Subjective Well-Being Measures. Journal of Public Economics (92): 1833–1845. Lelkes, O., 2006a. Knowing what is good for you: empirical analysis of personal preferences and the ‘‘objective good. The Journal of Socio-Economics 35, 285–307. Lelkes, O., 2006b. Tasting freedom: happiness, religion and economic transition. Journal of Economic Behavior and Organization 59, 173–194. Lukiyanova, Anna and Aleksey Oschepkov (2012) Income Mobility in Russia (2000–2005) Economic Systems (36): 46–64. Luttmer, Erzo (2005) Neighbors as Negatives: Relative Earnings and Well-Being. The Quarterly Journal of Economics (120, No. 3): 963–1002 Senik, Claudia (2005) Income Distribution and Well-Being: What Can We Learn From Subjective Data? Journal of Economic Surveys (19, No. 1): 43–63. Senik, Claudia (2009) Direct Evidence on Income Comparisons and their Welfare Effects. Journal of Economic Behavior & Organization (72): 408–424. Selezneva, Ekaterina (2011) Surveying Transitional Experience and Subjective Well-being: Income, Work, Family. Economic Systems (35): 139–157. Stiglitz, Joseph, Amartya Sen and Jean-Paul Fitoussi (2009) The Measurement of Economic Performance and Social Progress Revisited - Reflections and Overview. Commission on the Measurement of Economic Performance and Social Progress. Available from: http://www.stiglitz-sen-fitoussi.fr/documents/overview- eng.pdf Tortia, Ermanno C. (2008) Worker well-being and perceived fairness: Survey-based findings from Italy The Journal of Socio-Economics 37 (2080–2094) 25 | P a g e     Wolfers, J. (2003). Is business cycle volatility costly? Evidence from surveys of subjective well-being. International Finance, 6(1), 1–26. World Bank (2005) World Development Report 2006: Equity and Development. The World Bank and Oxford University Press ——— (2013a) World Development Indicators. [Data file] Available from: http://data.worldbank.org/data- catalog/world-development-indicators ——— (2013b). POVCALNET. [Data file] Available from: http://iresearch.worldbank.org/PovcalNet/index.htm?0 Zaidi, Salman; Alam, Asad; Mitra, Pradeep and Ramya Sundaram (2009) Satisfaction with Life and Service Delivery in Eastern Europe and the Former Soviet Union: Some Insights from the 2006 Life in Transition Survey. World Bank Working Paper 162. The World Bank, Washington, DC.Appendices 26 | P a g e     Appendices A1. Sample Characteristics Table A1.1: Countries included Region Countries Low/Lower-Middle Income Armenia Georgia Kyrgyzstan Moldova Countries (7) Tajikistan Ukraine Uzbekistan Albania Azerbaijan Belarus Bosnia and Herz. Bulgaria Croatia Czech Rep. Estonia Upper-Middle Income & Hungary Kazakhstan Latvia Lithuania Other Countries (21) Macedonia, FYR Montenegro Poland Romania Russian Fed. Serbia Slovakia Slovenia Turkey New EU Member States & Bulgaria Croatia Czech Rep. Estonia Turkey (12) Hungary Latvia Lithuania Poland Romania Slovakia Slovenia Turkey Western Balkans (5) Albania Bosnia and Herz. Macedonia, FYR Montenegro Serbia South Caucasus (3) Armenia Azerbaijan Georgia Central Asia (4) Kazakhstan Kyrgyzstan Tajikistan Uzbekistan Other (4) Moldova Ukraine Belarus Russian Fed. 27 | P a g e     Table A1.2: Perceived Movements in the Income Distribution 1989-2006 Countries Move  Down Same  Move Up New EU Member States Bulgaria 58.6 28.4 13.0 Croatia 56.0 35.4 8.6 Estonia 42.2 45.3 12.6 Czech Republic 28.1 52.4 19.4 Hungary 43.8 49.6 6.6 Latvia 49.0 38.4 12.6 Lithuania 43.4 43.7 12.9 Poland 40.0 45.5 14.5 Romania 39.9 49.4 10.8 Slovakia 37.7 51.5 10.9 Slovenia 24.4 62.1 13.5 Turkey 34.0 50.3 15.7 Western Balkans Albania 18.6 50.8 30.6 Bosnia and Herzegovina 68.1 24.3 7.6 Macedonia, FYR 60.0 34.0 6.0 Montenegro 65.8 27.9 6.3 Serbia 69.5 23.1 7.4 South Caucasus Armenia 53.2 31.0 15.9 Azerbaijan 64.0 29.1 6.9 Georgia 74.0 19.2 6.8 Central Asia Kazakhstan 40.7 42.2 17.1 Kyrgyzstan 46.6 40.0 13.4 Tajikistan 56.1 32.5 11.5 Uzbekistan 50.1 43.0 6.9 Russian Federation 52.1 34.7 13.2 Other Belarus 22.1 59.6 18.3 Moldova 42.0 46.6 11.4 Ukraine 51.3 37.9 10.9 Total 46.5 40.6 12.8 Source: Life in Transition Survey (2006) 28 | P a g e     Figure A1.1: GDP per capita, Gini and Public Employment Share Evolution 1989 - 2006 (a) GDP per cap. 1990 and lowest point 1990-2006 (b) GDP per cap. 1990 and change to 1990-2006 SVN SVN 8 8 1990 GDP per capita (2000 US$, thousands) 1990 GDP per capita (2000 US$, thousands) 6 6 CZE HRV SVK HRV CZE SVK LTU HUN LTUHUN 4 LVA 4 LVA TUR TUR POL POL RUS RUS MKD MKD 2 2 GEO ROM BGR GEO ROM SRB KAZ SRB BGR KAZ UKR AZE BLR UKR AZE BLR MDA ARM ALB MDA ALB ARM KGZ UZB UZB TJK TJK KGZ 0 0 20 40 60 80 100 50 100 150 200 Lowest GDP pc 1990-2006 as percentage of GDP pc 1990 2006 GDP per capita as percentage of GDP pc 1990 (c) Gini c. 1990 and change c.1990-c.2006 (d) Share of Public Empl. 1989 and change 1989-2006 45 90 BLR TUR UZB KAZ ALB GEO 40 80 BGR KGZ ARM AZE Share of Public Employment 1989 LTU SRB BIH MDA RUS UKR MNE LVA TJK Gini coefficient c. 1989 ROM MKD HRV 35 70 POL EST SVK HUN SVN 60 30 CZE POL KAZ KGZ 50 25 HUN UZB BGR SVN MDA RUS UKR HRV BLR ROM EST LVA LTU 40 20 SVK TUR -5 0 5 10 15 20 -50 -40 -30 -20 -10 Gini coefficient change c.1989 - c.2006 Change in Public Employment 1989 - 2006 Source: Panels (a), (b) and (c), WDI. Panel (d), 2006 LiTS. Figure A1.2: Share of time spend in Public/Private/Unemployed-Out of Labor Force by age in 2006 29 | P a g e     Figure A1.3. Household Perceived Position in 10-Step Income Ladder 2006 and 2010 (LiTS 2010) (a) 2006 (b) 2010 Figure A1.4. Perception of Economic Mobility 2006 – 2010 (a) ECA (b) New EU Member States (c) Western Balkans and Turkey ECA Type of Inter-step Mobility 2006 to 2010 EU10+1 Type of Inter-step Mobility 2006 to 2010 Western Balkans Type of Inter-step Mobility 2006 to 2010 80 80 80 76.2 73.3 69.5 60 60 60 Percentage Percentage Percentage 40 40 40 22.1 19.7 20 20 20 15.6 8.2 7.1 8.4 0 0 0 Down Same Up Down Same Up Down Same Up (d) Russia, Ukraine, Moldova (e) South Caucasus (f) Central Asia and Belarus CIS Type of Inter-step Mobility 2006 to 2010 South Caucasus Type of Inter-step Mobility 2006 to 2010 Central Asia Type of Inter-step Mobility 2006 to 2010 80.1 80 80 80 74.3 68.9 60 60 60 Percentage Percentage Percentage 40 40 40 18.5 20 20 20 14.0 14.6 12.6 11.0 5.9 0 0 0 Down Same Up Down Same Up Down Same Up Source: 2010 LiTS (individual weights) 30 | P a g e     Table A1.3: Sample estimation: Summary Statistics Standard  Share  / Mean Minimum Maximum Deviation Type of Perceived Economic Mobility Down 0.405 0 1 Same 0.438 0 1 Up 0.156 0 1 National Level Variables GDP per capita 1990 (2000 US$, thousands) 2.506 1.206 0.426 8.362 Change  GDP per capita 1990‐2006 (perc.) 0.182 0.322 ‐ 0.577 0.791 (Lowest GDPpc 1990‐2006)/GDPpc 1990 0.678 0.194 0.272 0.990 Gini  circa 1989 26.899 6.720 19.400 44.420 Change  in Gini  circa 1989‐2006 9.470 7.408 ‐11.580 18.330 Share  of Public Employment 1989 0.712 0.134 0.376 0.909 Change  on Public Employment 1989 ‐0.339 0.086 ‐ 0.508 ‐0.127 Household Characteristics Household Consumption Decile  (within country) Poorest 0.080 0 1 2 0.085 0 1 3 0.087 0 1 4 0.097 0 1 5 0.102 0 1 6 0.105 0 1 7 0.113 0 1 8 0.110 0 1 9 0.107 0 1 Richest 0.116 0 1 Location Rural 0.349 0 1 Urban 0.435 0 1 Metropolitan 0.216 0 1 Household Size 2.667 1.521 1 12 Share  Children <15 11.786 19.327 0 85.71 Share  Seniors 66+ 12.821 30.193 0 100 Household Receives Pension 0.459 0 1 Note: Countries not included: Bosnia and Herzegovina, Estonia and Montenegro (1990 GDP not available) (continued) 31 | P a g e     (continued) Standard  Share  / Mean Minimum Maximum Deviation Respondent Characteristics Age  in 2006 18‐34 0.298 0 1 35‐54 0.411 0 1 55‐65 0.169 0 1 66+ 0.122 0 1 Gender  0.484 0 1 Education No formal  education/Primary 0.186 0 1 Secondary 0.253 0 1 Professional/Vocational 0.318 0 1 University/Post‐graduate 0.243 0 1 Father's Education No formal  education/Primary 0.496 0 1 Secondary 0.195 0 1 Professional/Vocational 0.202 0 1 University/Post‐graduate 0.107 0 1 Ethnic Minority 0.078 0 1 Household head 0.691 0 1 Non‐ religious 0.108 0 1 Labor Status Out of Labor Force 0.395 0 1 Unemployment 0.053 0 1 Public Sector 0.212 0 1 Private Sector 0.256 0 1 Self Employed 0.087 0 1 Works in Agricultural  Sector 0.056 0 1 Employment History 1989‐ 2006 Share  of time  in Public Sector (wrt 18‐65) 0.315 0.379 0 1 Share  of time  in Private  Sector (wrt 18‐65) 0.282 0.368 0 1 Number of  Different Jobs 1.360 1.155 0 16 Subjective  Variables Mean Trust Indicator in Public Inst. (5 Institutions) ‐0.072 0.705 ‐1 1 Factor Most Important to Succeed Criminal  / Corrupts Ties / Political  Connections 0.305 Effort and Hard Work / Intellingence  and Skills 0.695 0 1 Support Market Economy (aot Planned/Indiff.) 0.402 0 1 Agree: Better than 1989 colleagues 0.219 0 1 Observations 17,325 Countries 25 Note: Countries not included: Bosnia and Herzegovina, Estonia and Montenegro (1990 GDP not available) 32 | P a g e     A2. Regression methodology The empirical specification we use for testing the relevance of the listed variables is as follows PM Cγ X β S φ u (1) where PM refers to perceived mobility, C to country-level variables, X to individual and household objective variables, S to individual subjective variables, and u to the error term. The sub-index i refers to individual and the sub-index j is added to make explicit that country variables are common for all individuals in a country. The use of subjective variables on both sides of the equation is not out of controversy. Hamermesh (2004) points to the fact that that if the subjective variables share a common error term, because of idiosyncrasies of the respondent, then any correlation obtained may be driven by this shared error term, rather by the actual experienced perception. More formally, if equation (1) were only composed by subjective variables, then we may have a problem if what is behind the observed variables is as follows: PM ∗ ε S∗ ε φ u (2) It is worth it to notice that the variables behind the observed subjective variables (PM* and S*) are not objective variables like actual income or age, but objective variables in the sense that can be measured unequivocally if we could measure the brain waves that occur as a reaction to an impulse. (Frey and Stutzer, 2002) Those are still perceptions, but not affected by the filter that memory, optimism, or other personal biases may impose in the declaration. The strategy we follow to avoid this problem is to extract information from PM that will most likely be driven by the actual perceived experience, rather than by the error term (ε ). For this, the variables we use for PM is the difference between two reported answers (step in 2006 and 1989). If they both share a common error term, then they both cancel out, and we have a clean measure of changes in perception. However, it can be argued that the error term in the declared current step may be different than the error term in the declared past step. If they both point in the same direction, taking first differences will diminish the magnitude of the distortion. However, if they point in opposite directions, the distortion may increase. Given this uncertainty, we decided to consider only changes that were of at least 2 steps. This way, we count as changes only the ones most likely driven not only by the error term, but by the actual perception. Certainly, we may be considering as not changes cases where there was an authentic change in the perception from 1989 to 2006, but we prefer to err in this direction than to consider as changes cases were is very likely are not driven by the actual perception. Hence, expressing explicitly the error term in the subjective variables, we claim that (1) can be expressed as: PM ∗ Cγ X β S∗ ε φ u (1’) This specification is more familiar in economics. Besides the common problem of possible endogeneity of C, X or S*, and attenuation of the parameter φ, another potential problem is that the error term in the subjective variable of the error term can be correlated with the error term of the equation. We have not a direct way to control for this problem in the estimation, but we performed robustness checks by estimating the same specification in different subsamples and we changed the subjective variables for similar ones that refer to different topics, but may arguably contain similar error terms. We estimated equation (1) using a probit model, where the category Moving Down is zero and, Same Step and Moving Up were considered as one. We decided to use this model to facilitate the interpretation of the 33 | P a g e     parameters. Also, the grouping in these two categories was decided because of the size of the groups, and because in a context where the average household declared having moved down, keeping the same relative position can be considered a success. Regarding other models we could have used, a multinomial logit using the same variables is presented in Table A2.5. This model shows results entirely consistent with the ones presented below. An ordered probit/logit was not considered either because of the difficulty to interpret the value of the parameters or marginal effects. The results are presented in the Tables A2.1 and following. We report the marginal effects to facilitate the interpretation of the effects. Also, we report only the significant variables obtained. Table A2.1: OLS Regression Results of Income and Perceived Mobility Dependent Variable: Change  in Steps 1989 ‐2006 ECA  (Expenditure) ECA  (Asset Index) OLS (1) (2) Ln(Monthly per capita  expenditure) 0.380*** (0.062) Asset Index 0.003*** (0.000) Constant ‐1.316*** 0.443** (0.345) (0.206) N 25270 25270 R^2 0.0539 0.0773 * p<0.10, ** p<0.05, *** p<0.01.  Dummy variables  by country omitted. Table A2.2: Probit Regression Results at Country Level (Marginal Effects) ECA (Country  ECA (Income  & Country variables) Dependent Variable: 1=Same  or moved up, 0=Moved down variables) Reporting Marginal Effects over Probability (1) (2) (3) (4) National  Level  Variables Change  GDP per capita  1990 ‐2006  (perc.) 0.199*** 0.197*** 0.038 0.06 (0.046) (0.046) (0.109) (0.044) GDP per capita  1990  (2000  US$, thousands) ‐0.026*** ‐0.026*** 0.008 0 (0.009) (0.009) (0.012) (0.015) Lowest GDPpc  1990 ‐2006)/GDPpc  1990 0.017 0.025 0.172 0.14 (0.098) (0.099) (0.229) (0.094) Change  in Gini  circa  1989 ‐2006 ‐0.001 ‐0.001 ‐0.005* (0.003) (0.003) (0.003) Gini  circa  1989 ‐0.013*** ‐0.014*** ‐0.080* (0.003) (0.003) (0.043) Changes  c.1989 ‐c.2006  ratio 80th /20th percentile ‐0.015* (0.009) Ratio 80th /20th percentile  (c.1989) 0.024 (0.024) Change  on Public  Employment 1989 0.546*** 0.555*** 0.470*** 0.369*** (0.085) (0.085) (0.107) (0.084) Share  of Public  Employment 1989 ‐0.574*** ‐0.557*** 0.514 ‐0.085 (0.166) (0.166) (0.376) (0.168) HH Consumption decile X X X * p<0.10, ** p<0.05, *** p<0.01.  Countries  not included: Bosnia  and Herzegovina, Estonia  and Montenegro (1990  GDP not available) in (1) &  (2); Albania, Armenia, Azerbaijan, Bosnia   and Herzegovina, Czech Republic, Estonia, Macedonia  (FYR), Georgia, Montenegro, Serbia, Tajikistan, Uzbekistan, in (3); Bosnia  and Herzegovina,  Croatia, Estonia, FYROM, Montenegro, Serbia  and Slovenia, in (4) 34 | P a g e     Table A2.3: Probit Regression Results (Marginal Effects) ECA  (Objective  &  Low/Lower‐Middle   Upper‐Middle  Income    ECA  (All Variables) Dependent Variable: 1=Same  or moved up, 0=Moved down Country variables) Income  Countries & Other Countries Reporting Marginal Effects over Probability (1) (2) (3) (4) National  Level  Variables GDP  per capita  1990  (2000  US$, thousands) ‐0.025*** ‐0.030*** (0.009) (0.009) Change  GDP  per capita  1990 ‐2006  (perc.) 0.213*** 0.279*** (0.044) (0.043) (Lowest  GDPpc  1990 ‐2006)/GDPpc  1990 0.011 ‐0.191** (0.092) (0.093) Gini  circa  1989 ‐0.014*** ‐0.017*** (0.003) (0.003) Change  in Gini  circa  1989 ‐2006 ‐0.006 ‐0.008 (0.031) (0.030) Share  of Public  Employment 1989 ‐0.645*** ‐0.911*** (0.167) (0.161) Change  on Public  Employment  1989 0.418*** 0.292*** (0.085) (0.088) Respondent Age Age  18  ‐ 34 0.072*** 0.073*** 0.153*** 0.050** (0.021) (0.022) (0.038) (0.025) Age  35  ‐ 54  (Base) Age  54  ‐ 65 0.031 0.042 0.046 0.04 (0.029) (0.029) (0.041) (0.031) Age  66+ 0.026 0.053 ‐0.016 0.075 (0.050) (0.048) (0.067) (0.054) Respondent Education Primary or less  (Base) Secondary 0.002 0.01 0.043 0.006 (0.025) (0.025) (0.045) (0.028) Professional/Vocational ‐0.025 ‐0.016 ‐0.009 ‐0.001 (0.026) (0.026) (0.045) (0.028) University/Post ‐graduate 0.054* 0.059* 0.023 0.077** (0.030) (0.031) (0.051) (0.034) Respondent's  Father Education Primary or less  (Base) Secondary 0.068*** 0.060** ‐0.027 0.095*** (0.024) (0.024) (0.032) (0.030) Professional/Vocational 0.075*** 0.075*** 0.014 0.094*** (0.028) (0.029) (0.045) (0.033) University/Post ‐graduate 0.036 0.029 0.083* 0.01 (0.035) (0.036) (0.047) (0.043) Respondent Labor Status Out of Labor Force  (base) Unemployed ‐0.051 ‐0.037 ‐0.064 ‐0.033 (0.035) (0.033) (0.048) (0.039) Employed PUBLIC 0.041 0.053* 0.03 0.056 (0.029) (0.030) (0.046) (0.036) Employed PRIVATE 0.029 0.034 0.008 0.038 (0.030) (0.030) (0.054) (0.032) Employed SELF‐EMPLOYED 0.072** 0.066** 0.085* 0.054 (0.031) (0.032) (0.050) (0.040) Employed Agricultural  Sector ‐0.045 ‐0.057 ‐0.032 ‐0.059 (0.035) (0.035) (0.052) (0.046) Share  of time  employed PUBLIC (wrt  18 ‐65) ‐0.060* ‐0.056* ‐0.064 ‐0.055 (0.033) (0.034) (0.062) (0.037) Share  of time  employed PRIVATE (wrt 18 ‐65) ‐0.004 0.003 ‐0.056 0.012 (0.034) (0.036) (0.051) (0.040) Number of Jobs ‐0.003 ‐0.005 ‐0.015 ‐0.005 (0.011) (0.011) (0.017) (0.012) Mean Trust  Public  Institutions  (5) 0.068*** 0.101*** 0.060*** (0.018) (0.024) (0.021) Factor Most  Important to Succeed Criminal  / Corrupts  Ties  / Political  Connections  (Base) Effort  and Hard Work / Intellingence  and Skills 0.086*** 0.05 0.086*** (0.018) (0.038) (0.019) Support Market  Economy (aot  Planned/Indiff.) 0.073*** 0.069** 0.067*** (0.018) (0.033) (0.020) Better than 1989  Colleagues  (aot  Worse/Indiff.) 0.037* 0.101*** 0.019 (0.022) (0.033) (0.025) HH Consumption decile X X X X Country dummies X X * p<0.10, ** p<0.05, *** p<0.01.  Variables  included but not reported: (Household) Urban/Rural/Metropolitan, size, share  children <15, share  senior 66+, household received pensions,  (Respondent) gender, ethnic  minority, household head, non‐religious.  Countries  not included in (1) and (2): Bosnia  and Herzegovina, Estonia  and Montenegro (1990  GDP not available) 35 | P a g e     Table A2.4: Probit Regression Results by Sub-Regions (Marginal Effects) New EU  Ukraine,  Member  Western  Moldova and  South Caucasus Central Asia Russia States and  Balkans 1 Belarus Dependent Variable: 1=Same  or moved up, 0=Moved down Turkey  Reporting Marginal Effects over Probability (1) (2) (3) (4) (5) (6) Respondent Age Age  18  ‐ 34 0.029 0.062* 0.140** 0.058* 0.074** 0.094 (0.023) (0.034) (0.057) (0.033) (0.034) (0.070) Age  35  ‐ 54  (Base) Age  54  ‐ 65 0.003 ‐0.104*** 0.111** 0.05 ‐0.074 0.116 (0.025) (0.036) (0.044) (0.047) (0.047) (0.078) Age  66+ 0.057 ‐0.139*** ‐0.014 0.05 0.086 0.152 (0.047) (0.049) (0.101) (0.054) (0.070) (0.139) Respondent Education Primary or less  (Base) Secondary 0.062*** ‐0.128*** 0.054 ‐0.087** 0.094* ‐0.116 (0.023) (0.033) (0.055) (0.043) (0.052) (0.085) Professional/Vocational ‐0.003 ‐0.151*** ‐0.016 ‐0.067 0.061 ‐0.034 (0.026) (0.033) (0.052) (0.050) (0.053) (0.082) University/Post‐graduate 0.062* ‐0.087** 0.031 ‐0.097* 0.109* 0.049 (0.036) (0.038) (0.062) (0.053) (0.057) (0.085) Respondent Father's  Education Primary or less  (Base) Secondary 0.056** ‐0.027 ‐0.014 ‐0.005 0.004 0.158** (0.027) (0.042) (0.055) (0.037) (0.030) (0.059) Professional/Vocational 0.038 ‐0.005 0.058 0.06 0 0.156** (0.024) (0.038) (0.055) (0.054) (0.044) (0.071) University/Post‐graduate 0.016 ‐0.003 0.131** ‐0.021 ‐0.058 0.035 (0.038) (0.058) (0.059) (0.048) (0.055) (0.075) Respondent Labor Status Out of Labor Force  (base) Unemployed ‐0.033 ‐0.075* ‐0.066 ‐0.03 ‐0.027 ‐0.081 (0.040) (0.039) (0.072) (0.043) (0.061) (0.104) Employed PUBLIC 0.078** ‐0.070* 0.023 0.112** 0.033 0.03 (0.033) (0.042) (0.060) (0.057) (0.043) (0.079) Employed PRIVATE 0.01 ‐0.005 ‐0.033 0.032 0.128*** 0.028 (0.028) (0.042) (0.064) (0.063) (0.049) (0.077) Employed SELF‐EMPLOYED ‐0.014 0.057 0.159** ‐0.032 0.083* 0.324* (0.040) (0.050) (0.076) (0.069) (0.046) (0.170) Employed Agricultural  Sector ‐0.013 ‐0.052 ‐0.04 0.034 ‐0.077* ‐0.126 (0.056) (0.071) (0.085) (0.083) (0.045) (0.152) Share  of time  employed PUBLIC (wrt 18 ‐65) ‐0.059* 0.039 ‐0.125 ‐0.092 0.101* ‐0.005 (0.036) (0.055) (0.085) (0.066) (0.055) (0.095) Share  of time  employed PRIVATE (wrt 18 ‐65) ‐0.013 ‐0.029 ‐0.086 ‐0.094 0.066 0.137 (0.034) (0.050) (0.067) (0.071) (0.061) (0.110) Number of Jobs ‐0.011 ‐0.015 ‐0.024 ‐0.001 ‐0.038*** 0.003 (0.008) (0.016) (0.023) (0.024) (0.014) (0.025) Mean Trust Public  Institutions  (5) 0.084*** 0.092*** 0.090*** 0.063*** 0.088*** 0.027 (0.017) (0.017) (0.028) (0.023) (0.023) (0.038) Factor Most Important to Succeed Criminal  / Corrupts  Ties  / Political  Connections  (Base) Effort and Hard Work / Intellingence  and Skills 0.037** 0.068*** 0.027 0.077** 0.059 0.156*** (0.018) (0.025) (0.041) (0.031) (0.036) (0.038) Support  Market Economy (aot Planned/Indiff.) 0.069*** ‐0.023 0.071* 0.043 0.081*** 0.051 (0.020) (0.025) (0.042) (0.029) (0.031) (0.050) Better than 1989  Colleagues  (aot Worse/Indiff.) 0.011 0.034 0.046 0.138*** 0.132*** 0.012 (0.021) (0.026) (0.043) (0.044) (0.034) (0.059) Country dummies X X X X X X HH Consumption decile X X X X X X * p<0.10, ** p<0.05, *** p<0.01.  Variables  included but not reported: (Household) Urban/Rural/Metropolitan, size, share  children <15, share  senior 66+,  household received pensions, (Respondent) gender, ethnic  minority, household head, non‐religious.  New EU Member States  include  Bulgaria, Croatia,  Czech Republic, Hungary, Latvia, Lithuania,  Poland, Romania, Slovakia, and Slovenia. 36 | P a g e     Table A2.5: Multinomial Logit Regression (Marginal Effects) Dependent Variable: Type  of movement ECA ‐ Down ECA  ‐ Same ECA ‐ Up Reporting Marginal Effects over Probability (1) (2) (3) National  Level  Variables GDP  per capita  1990  (2000  US$, thousands) 0.030*** ‐0.016* ‐0.013 (0.009) (0.009) (0.008) Change  GDP  per capita  1990 ‐2006  (perc.) ‐0.271*** 0.036 0.235*** (0.042) (0.048) (0.038) (Lowest  GDPpc  1990 ‐2006)/GDPpc  1990 0.193** 0.254*** ‐0.447*** (0.091) (0.093) (0.095) Gini  circa  1989 0.016*** ‐0.012*** ‐0.004 (0.003) (0.003) (0.003) Change  in Gini  circa  1989 ‐2006 0.006 ‐0.049 0.043 (0.030) (0.035) (0.034) Share  of Public  Employment  1989 0.909*** ‐0.300* ‐0.609*** (0.155) (0.172) (0.168) Change  on Public  Employment 1989 ‐0.267*** 0.325*** ‐0.059 (0.086) (0.094) (0.087) Respondent  Age Age  18  ‐ 34 ‐0.073*** 0.054** 0.019 (0.022) (0.022) (0.023) Age  35  ‐ 54  (Base) Age  54  ‐ 65 ‐0.04 0.073** ‐0.033 (0.027) (0.033) (0.022) Age  66+ ‐0.052 0.053 ‐0.001 (0.046) (0.058) (0.053) Respondent  Education Primary or less  (Base) Secondary ‐0.009 0.019 ‐0.01 (0.024) (0.028) (0.024) Professional/Vocational 0.016 0.003 ‐0.019 (0.025) (0.031) (0.025) University/Post ‐graduate ‐0.055* 0.044 0.011 (0.029) (0.035) (0.027) Respondent  Father's  Education Primary or less  (Base) Secondary ‐0.057** 0.086*** ‐0.029 (0.024) (0.026) (0.019) Professional/Vocational ‐0.074** 0.075** ‐0.001 (0.029) (0.030) (0.023) University/Post ‐graduate ‐0.028 0.069 ‐0.041 (0.036) (0.045) (0.027) Respondent  Labor Status Out of Labor Force  (base) Unemployed 0.036 ‐0.036 0 (0.033) (0.034) (0.027) Employed PUBLIC ‐0.053* 0.006 0.047* (0.028) (0.031) (0.027) Employed PRIVATE ‐0.035 0.01 0.025 (0.029) (0.028) (0.024) Employed SELF‐EMPLOYED ‐0.065** 0.019 0.046 (0.031) (0.038) (0.037) Employed Agricultural  Sector 0.057 ‐0.021 ‐0.035 (0.035) (0.039) (0.032) Share  of time  employed PUBLIC (wrt 18 ‐65) 0.055 ‐0.022 ‐0.034 (0.034) (0.037) (0.033) Share  of time  employed PRIVATE (wrt  18 ‐65) ‐0.003 ‐0.009 0.012 (0.036) (0.034) (0.030) Number of Jobs 0.005 ‐0.003 ‐0.002 (0.011) (0.011) (0.007) Mean Trust Public  Institutions  (5) ‐0.065*** 0.038** 0.027** (0.018) (0.016) (0.013) Factor Most Important  to Succeed Criminal  / Corrupts  Ties  / Political  Connections  (Base) Effort and Hard Work / Intellingence  and Skills ‐0.084*** 0.031 0.053*** (0.017) (0.021) (0.017) Support  Market  Economy (aot Planned/Indiff.) ‐0.070*** 0.047** 0.023 (0.018) (0.019) (0.019) Better than 1989  Colleagues  (aot  Worse/Indiff.) ‐0.036* ‐0.022 0.058** (0.022) (0.020) (0.024) HH Consumption decile X X X * p<0.10, ** p<0.05, *** p<0.01.  Variables  included but  not  reported: (Household) Urban/Rural/Metropolitan, size, share  children <15, share  senior 66+,  household received pensions, (Respondent) gender, ethnic  minority, household head, non‐religious.  Countries  not included: Bosnia  and Herzegovina, Estonia  and Montenegro (1990  GDP  not available) 37 | P a g e     A3. Robustness checks To check the consistency of the parameters, we performed some robustness checks. We first check how sensitive our parameters are to our definition of the dependent variables (Columns (1) and (2)). National and subjective variables show to be very consistent. The individual level variables show some more significant results when the changes in one step are counted, and some lose significance when moving to a three-step definition, but neither age, nor education nor employment status show sign changes or dramatic magnitude changes. Next, we wanted to control for the fact that upper steps are more likely to move down, and the opposite for the lower steps, if we assume that they have equal chances to move to any different step. If this is the case, then the parameters may be just capturing the characteristics of households in top upper or lower steps to identify the effects. This exercise is besides excluding the upper and lower 2006 step to consider only households that can move in any direction. We estimate the regression using different samples, that included the central steps (Columns (4) and (5)), and the extreme steps (Columns (6) and (7)), where this bias is more likely to be present. National level and subjective variables show to be consistent across different samples, with the exception of comparison with local groups, which is not significant except for the initial specification. At the individual level, young age shows to be very consistent across samples. Higher education loses significance, but it always has a positive sign. Self-employment is only significant in the whole sample, and shows some sign changes. We explored also how likely is the significance of the subjective variables being driven by the error term contained being correlated with the dependent or the error term of (1), and we obtain that most likely is not the case for trust in public institutions, perceptions of fairness and support to market economy. However, for comparison with local groups, results suggest the result may be driven by individual idiosyncrasies. Our strategy consisted in including other subjective variables that can arguably contain a similar error term, but depend on different perceptions. We included complete trust in religious institutions (for trust in public institutions), support to democracy (for support to market economy), and perceptions of the country being doing better than in 1989 (for comparison with local groups). Results in column (2) of Table A3.2 show that the first two ones are not significant, which suggest that the correlation observed in the original variables may not be driven by the error term of the subjective variable. The last variable, the country doing better than in 1989, was significant. A second exercise we performed was to restrict the sample only to the cases where the respondent considered their household doing worse than in 1989, in absolute terms. Within this sample, the error of the specification (1) will very likely not be associated to unobserved factors like optimistic or pessimistic biases influencing the dependent variable, but we can assume more confidently it has a more random distribution. The results in (3) show that the subjective variables are still significant, with the exception the comparison to local groups. This result suggests that the observed correlation in the whole sample may be driven by correlation of the error term in this subjective variable with the error term of (1). Column (4) show the results with the other part of the sample, households doing same or equal than in 1989. National level variables are less significant, but still show the same sign and similar magnitudes. Subjective variables are less significant, and again comparison to local groups is not significant. In this case, also support to market economy is not significant. Finally, given that in our estimation sample we included young workers who were under 18 years old in 1989, we wanted to confirm whether results hold if we consider only workers older than 18 in 1989. For this, we first restricted the sample to the cases when the respondent is the same to the perception and employment section. 38 | P a g e     Then, we restricted even further, by considering only workers 35 or older in 2006, and that responded both sections of the survey. Results are presented in columns (5) and (6) of Table A3.2, and show results entirely consistent with the whole estimation sample. The only noticeable difference is in the significance of belonging to the 55-64 age group. The reported negative effect over the probabilities of perceiving downward mobility turns statistically significant, though only weakly. Also, the share of time spent working in the public sector becomes significant, showing now that more time in the public sector between 1989 and 2006 is associated with more probabilities of perceiving a movement down. All the other parameters sign and significance remain unaltered. 39 | P a g e     Table A3.1 Changes 1  step Change  2  steps Change  3  steps Steps 5 & 6 Steps 7 & 8 Steps 3  & 4 Dependent Variable: 1=Same  or moved up, 0=Moved down Reporting Marginal Effects over Probability (1) (2) (3) (4) (6) (7) National  Level  Variables GDP  per capita  1990  (2000 US$, thousands) ‐0.024*** ‐0.030*** ‐0.010* ‐0.031*** ‐0.013* ‐0.033** (0.008) (0.009) (0.006) (0.010) (0.008) (0.013) Change  GDP  per capita  1990 ‐2006  (perc.) 0.308*** 0.279*** 0.099*** 0.256*** 0.013 0.335*** (0.037) (0.043) (0.024) (0.049) (0.035) (0.064) (Lowest  GDPpc  1990 ‐2006)/GDPpc 1990 ‐0.467*** ‐0.191** ‐0.157*** ‐0.261** 0.032 ‐0.303** (0.086) (0.093) (0.059) (0.109) (0.075) (0.126) Gini  circa  1989 ‐0.009*** ‐0.017*** ‐0.002 ‐0.018*** ‐0.004* ‐0.014*** (0.003) (0.003) (0.002) (0.004) (0.002) (0.004) Change  in Gini  circa  1989 ‐2006 0.035 ‐0.008 0.009 ‐0.022 ‐0.016 0.043 (0.029) (0.030) (0.018) (0.042) (0.018) (0.041) Share  of Public Employment  1989 ‐0.721*** ‐0.911*** ‐0.311*** ‐0.830*** ‐0.221* ‐1.089*** (0.159) (0.161) (0.114) (0.168) (0.128) (0.231) Change  on Public Employment  1989 0.05 0.292*** ‐0.099** 0.321*** 0.139* 0.279** (0.085) (0.088) (0.048) (0.110) (0.078) (0.132) Respondent  Age Age  18 ‐ 34 0.055** 0.073*** 0.016 0.042 0.061* 0.052 (0.024) (0.022) (0.013) (0.030) (0.035) (0.034) Age  35 ‐ 54  (Base) Age  54 ‐ 65 ‐0.026 0.042 ‐0.015 ‐0.007 ‐0.009 0.102** (0.021) (0.029) (0.011) (0.032) (0.023) (0.042) Age  66+ 0.003 0.053 ‐0.004 0.029 0.017 0.083 (0.047) (0.048) (0.021) (0.062) (0.024) (0.068) Respondent  Education Primary or less  (Base) Secondary 0.016 0.01 0.005 ‐0.002 0.003 0.002 (0.023) (0.025) (0.011) (0.031) (0.019) (0.040) Professional/Vocational 0.007 ‐0.016 0.028** ‐0.056* 0.007 ‐0.031 (0.022) (0.026) (0.014) (0.033) (0.020) (0.039) University/Post‐graduate 0.054* 0.059* 0.024 0.057 0.016 ‐0.004 (0.029) (0.031) (0.015) (0.039) (0.014) (0.050) Respondent  Father's  Education Primary or less  (Base) Secondary 0.004 0.060** ‐0.009 0.016 0.003 0.075* (0.019) (0.024) (0.012) (0.026) (0.012) (0.042) Professional/Vocational 0.027 0.075*** ‐0.004 0.008 0.011 0.123*** (0.023) (0.029) (0.014) (0.033) (0.015) (0.044) University/Post‐graduate 0.007 0.029 ‐0.019 ‐0.042 ‐0.049 0.110* (0.031) (0.036) (0.014) (0.044) (0.037) (0.060) Respondent  Labor Status Out  of Labor Force  (base) Unemployed ‐0.032 ‐0.037 ‐0.021 ‐0.013 ‐0.029 ‐0.016 (0.023) (0.033) (0.013) (0.046) (0.041) (0.046) Employed PUBLIC 0.066** 0.053* 0.007 0.016 ‐0.028 0.104** (0.028) (0.030) (0.017) (0.036) (0.032) (0.047) Employed PRIVATE 0.045* 0.034 ‐0.013 0.017 0 0.053 (0.024) (0.030) (0.012) (0.034) (0.018) (0.043) Employed SELF‐EMPLOYED 0.087*** 0.066** 0.008 0.063* ‐0.011 0.034 (0.029) (0.032) (0.019) (0.035) (0.024) (0.062) Employed Agricultural  Sector ‐0.067*** ‐0.057 ‐0.016 ‐0.065 0.021 ‐0.029 (0.024) (0.035) (0.021) (0.049) (0.013) (0.054) Share  of time  employed PUBLIC (wrt 18 ‐65) ‐0.045 ‐0.056* ‐0.005 ‐0.002 0 ‐0.069 (0.030) (0.034) (0.019) (0.045) (0.018) (0.055) Share  of time  employed PRIVATE (wrt 18 ‐65) ‐0.012 0.003 0.019 0.007 ‐0.021 0.044 (0.031) (0.036) (0.015) (0.046) (0.021) (0.051) Number of Jobs 0 ‐0.005 0.001 0.007 0.01 ‐0.014 (0.008) (0.011) (0.005) (0.014) (0.008) (0.012) Mean Trust  Public  Institutions  (5) 0.038*** 0.068*** 0.011* 0.035 0.017* 0.081*** (0.012) (0.018) (0.006) (0.022) (0.010) (0.023) Factor Most  Important  to Succeed Criminal  / Corrupts  Ties  / Political  Connections  (Base) Effort  and Hard Work / Intellingence  and Skills 0.048*** 0.086*** 0.017 0.055** 0.017 0.069** (0.015) (0.018) (0.010) (0.026) (0.015) (0.029) Support  Market Economy (aot  Planned/Indiff.) 0.042*** 0.073*** 0.022** 0.074*** 0.035** 0.040* (0.016) (0.018) (0.011) (0.025) (0.017) (0.024) Better than 1989  Colleagues  (aot  Worse/Indiff.) 0.082*** 0.037* 0.046*** ‐0.019 0.003 0.008 (0.024) (0.022) (0.014) (0.025) (0.009) (0.030) HH Consumption decile X X X X X X * p<0.10, ** p<0.05, *** p<0.01.  Variables  included but  not  reported: (Household) Urban/Rural/Metropolitan, size, share  children <15, share  senior 66+, household received pensions,   (Respondent) gender, ethnic  minority, household head, non‐religious.  Countries  not included: Bosnia  and Herzegovina, Estonia  and Montenegro (1990  GDP not  available) 40 | P a g e     Table A3.2 Alternative   Only HHs doing  Only HHs doing  ECA  (All, only HH  ECA (All, only HH  All Variables subjective   worse  than in  same/better than  Heads) Heads, 35+) Dependent Variable: 1=Same  or moved up, 0=Moved down variables 1989 in 1989 Reporting Marginal  Effects over Probability (1) (2) (3) (4) (5) (6) National  Level  Variables GDP per capita  1990  (2000  US$, thousands) ‐0.030*** ‐0.026**  ‐0.035*** ‐0.007 ‐0.031*** ‐0.028** (0.009) (0.010) (0.012) (0.009) (0.010) (0.011) Change  GDP per capita  1990 ‐2006  (perc.) 0.279*** 0.142*** 0.164*** 0.143*** 0.309*** 0.309*** (0.043) (0.047) (0.058) (0.054) (0.053) (0.058) (Lowest GDPpc  1990 ‐2006)/GDPpc  1990 ‐0.191** 0.061 0.075 ‐0.216** ‐0.274** ‐0.213* (0.093) (0.096) (0.119) (0.103) (0.114) (0.128) Gini  circa  1989 ‐0.017*** ‐0.016*** ‐0.018*** ‐0.007** ‐0.017*** ‐0.015*** (0.003) (0.003) (0.004) (0.003) (0.003) (0.004) Change  in Gini  circa  1989 ‐2006 ‐0.008 ‐0.041 ‐0.043 0.006 0.003 ‐0.006 (0.030) (0.034) (0.035) (0.033) (0.037) (0.036) Share  of Public  Employment 1989 ‐0.911*** ‐0.770*** ‐0.892*** ‐0.356** ‐0.903*** ‐0.829*** (0.161) (0.175) (0.218) (0.172) (0.185) (0.213) Change  on Public  Employment 1989 0.292*** 0.267*** 0.337*** 0.253*** 0.272*** 0.273** (0.088) (0.086) (0.118) (0.098) (0.103) (0.115) Respondent Age Age  18  ‐ 34 0.073*** 0.067*** 0.068** 0.065** 0.095*** (0.022) (0.022) (0.030) (0.027) (0.031) Age  35  ‐ 54  (Base) Age  54  ‐ 65 0.042 0.044 0.078** 0.034 0.060* 0.076** (0.029) (0.030) (0.036) (0.025) (0.032) (0.037) Age  66+ 0.053 0.048 0.068 0.061 0.092* 0.108* (0.048) (0.052) (0.060) (0.044) (0.055) (0.063) Respondent Education Primary or less  (Base) Secondary 0.01 0.014 0.019 ‐0.004 0.018 0.021 (0.025) (0.028) (0.033) (0.034) (0.032) (0.035) Professional/Vocational ‐0.016 ‐0.017 ‐0.002 ‐0.004 ‐0.005 0.005 (0.026) (0.027) (0.034) (0.037) (0.031) (0.036) University/Post‐graduate 0.059* 0.048 0.061 0.014 0.068* 0.084** (0.031) (0.034) (0.041) (0.037) (0.038) (0.042) Respondent Father's  Education Primary or less  (Base) Secondary 0.060** 0.070*** 0.060* 0.073*** 0.051* 0.062* (0.024) (0.025) (0.031) (0.028) (0.029) (0.033) Professional/Vocational 0.075*** 0.066**  0.064* 0.066* 0.063* 0.086** (0.029) (0.031) (0.037) (0.034) (0.034) (0.043) University/Post‐graduate 0.029 0.026 ‐0.007 0.072* 0.015 ‐0.064 (0.036) (0.037) (0.048) (0.041) (0.044) (0.059) Respondent Labor Status Out of Labor Force  (base) Unemployed ‐0.037 ‐0.043 ‐0.003 ‐0.024 ‐0.077 ‐0.099 (0.033) (0.038) (0.045) (0.048) (0.050) (0.062) Employed PUBLIC 0.053* 0.053*   0.056 0.057** 0.058* 0.079** (0.030) (0.030) (0.041) (0.023) (0.033) (0.040) Employed PRIVATE 0.034 0.027 0.029 0.01 0.03 0.035 (0.030) (0.033) (0.041) (0.034) (0.037) (0.046) Employed SELF‐EMPLOYED 0.066** 0.051 0.002 0.063** 0.059 0.08 (0.032) (0.036) (0.048) (0.030) (0.039) (0.050) Employed Agricultural  Sector ‐0.057 ‐0.033 ‐0.066 0.023 ‐0.083* ‐0.114** (0.035) (0.035) (0.046) (0.035) (0.047) (0.052) Share  of time  employed PUBLIC (wrt 18 ‐65) ‐0.056* ‐0.066*   ‐0.052 ‐0.029 ‐0.111*** ‐0.094** (0.034) (0.037) (0.044) (0.038) (0.041) (0.044) Share  of time  employed PRIVATE (wrt 18 ‐65) 0.003 ‐0.015 0.023 0.033 ‐0.024 0.014 (0.036) (0.034) (0.045) (0.042) (0.046) (0.048) Number of Jobs ‐0.005 ‐0.006 ‐0.005 ‐0.001 0.003 0.009 (0.011) (0.011) (0.014) (0.009) (0.014) (0.015) Mean  Trust Public  Institutions  (5) 0.068*** 0.056*** 0.038** 0.060*** 0.066*** (0.018) (0.019) (0.016) (0.023) (0.023) Factor Most Important to Succeed Criminal  / Corrupts  Ties  / Political  Connections  (Base) Effort and Hard Work / Intellingence  and Skills 0.086*** 0.084*** 0.033 0.095*** 0.128*** (0.018) (0.025) (0.025) (0.022) (0.026) Support  Market Economy (aot Planned/Indiff.) 0.073*** 0.053** 0.026 0.076*** 0.076*** (0.018) (0.024) (0.019) (0.021) (0.024) Better than 1989  Colleagues  (aot  Worse/Indiff.) 0.037* ‐0.028 ‐0.017 0.043 0.043 (0.022) (0.029) (0.023) (0.027) (0.030) Trust in Religious  Institutions 0.031 (0.024) Support  Democracy (aot Authoritarian/Indiff.) 0.028 (0.020) Economy country better than 1989  (aot Worse/Indiff.) 0.244*** (0.021) HH Consumption decile X X X X X X * p<0.10, ** p<0.05,  *** p<0.01.  Variables  included but not reported: (Household) Urban/Rural/Metropolitan, size, share  children <15, share  senior 66+, household received pensions,  (Respondent) gender, ethnic  minority,  household head, non ‐religious. Countries  not included: Bosnia  and Herzegovina, Estonia  and Montenegro (1990  GDP  not available) 41 | P a g e     A4. Perceived Movements by Individual Characteristics: Heads and Other Member Table A4.1 Perceived Movements by Age: Household Heads and Other Members 18-34 35-54 55-65 66+ HH HH HH HH All Other All Other All Other All Other Head Head Head Head Down 35.3 32.5 37.9 47.3 48.1 45.2 50.3 50.9 47.3 59.2 60.0 52.9 Same 47.5 47.6 47.5 38.7 38.1 40.3 40.4 40.0 42.5 34.3 34.2 35.2 Up 17.2 19.9 14.6 14.0 13.8 14.6 9.3 9.1 10.1 6.5 5.8 12.0 Figure A4.1 Perceived Movements by Age: Household Heads and Other Members (a) All Sample (b) Only Household Heads (Same respondent) (c) Other Member (Different respondents) 42 | P a g e     Table A4.2 Perceived Movements by Level of Education: Household Heads and Other Members No formal / Primary Secondary Professional/Vocational University/Post-graduate HH HH HH HH All Other All Other All Other All Other Head Head Head Head Down 46.7 50.2 38.8 47.5 50.5 42.4 50.2 50.9 48.3 39.8 40.1 38.7 Same 43.5 41.5 47.8 40.0 37.3 44.5 37.6 36.7 40.0 42.7 43.2 40.9 Up 9.8 8.3 13.4 12.5 12.2 13.2 12.2 12.4 11.8 17.5 16.7 20.3 Figure A4.2 Perceived Movements by Level of Education: Household Heads and Other Members (a) All Sample (b) Only Household Heads (Same respondent) (c) Other Member (Different respondents) 43 | P a g e     Table A4.3 Perceived Movements by Employment Status: Household Heads and Other Members Employed Unemployed Out of Labor Force HH HH HH All Other All Other All Other Head Head Head Down 42.1 41.9 42.5 52.4 55.5 47.7 51.2 55.5 55.5 Same 41.8 41.6 42.2 37.8 36.3 40.1 39.6 37.1 37.1 Up 16.1 16.5 15.3 9.8 8.3 12.2 9.1 7.4 7.4 Figure A4.3 Perceived Movements by Employment Status: Household Heads and Other Members (a) All Sample (b) Only Household Heads (Same respondent) (c) Other Member (Different respondents) 44 | P a g e     Poverty & Equity Global Practice Working Papers (Since July 2014) The Poverty & Equity Global Practice Working Paper Series disseminates the findings of work in progress to encourage the exchange of ideas about development issues. An objective of the series is to get the findings out quickly, even if the presentations are less than fully polished. The papers carry the names of the authors and should be cited accordingly. The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent the views of the International Bank for Reconstruction and Development/World Bank and its affiliated organizations, or those of the Executive Directors of the World Bank or the governments they represent. This series is co‐published with the World Bank Policy Research Working Papers (DECOS). It is part of a larger effort by the World Bank to provide open access to its research and contribute to development policy discussions around the world. For the latest paper, visit our GP’s intranet at http://POVERTY. 1 Estimating poverty in the absence of consumption data: the case of Liberia Dabalen, A. L., Graham, E., Himelein, K., Mungai, R., September 2014 2 Female labor participation in the Arab world: some evidence from panel data in Morocco Barry, A. G., Guennouni, J., Verme, P., September 2014 3 Should income inequality be reduced and who should benefit? redistributive preferences in Europe and Central Asia Cojocaru, A., Diagne, M. F., November 2014 4 Rent imputation for welfare measurement: a review of methodologies and empirical findings Balcazar Salazar, C. F., Ceriani, L., Olivieri, S., Ranzani, M., November 2014 5 Can agricultural households farm their way out of poverty? Oseni, G., McGee, K., Dabalen, A., November 2014 6 Durable goods and poverty measurement Amendola, N., Vecchi, G., November 2014 7 Inequality stagnation in Latin America in the aftermath of the global financial crisis Cord, L., Barriga Cabanillas, O., Lucchetti, L., Rodriguez‐Castelan, C., Sousa, L. D., Valderrama, D. December 2014 8 Born with a silver spoon: inequality in educational achievement across the world Balcazar Salazar, C. F., Narayan, A., Tiwari, S., January 2015 Updated on December 2015 by POV GP KL Team | 1 9 Long‐run effects of democracy on income inequality: evidence from repeated cross‐sections Balcazar Salazar,C. F., January 2015 10 Living on the edge: vulnerability to poverty and public transfers in Mexico Ortiz‐Juarez, E., Rodriguez‐Castelan, C., De La Fuente, A., January 2015 11 Moldova: a story of upward economic mobility Davalos, M. E., Meyer, M., January 2015 12 Broken gears: the value added of higher education on teachers' academic achievement Balcazar Salazar, C. F., Nopo, H., January 2015 13 Can we measure resilience? a proposed method and evidence from countries in the Sahel Alfani, F., Dabalen, A. L., Fisker, P., Molini, V., January 2015 14 Vulnerability to malnutrition in the West African Sahel Alfani, F., Dabalen, A. L., Fisker, P., Molini, V., January 2015 15 Economic mobility in Europe and Central Asia: exploring patterns and uncovering puzzles Cancho, C., Davalos, M. E., Demarchi, G., Meyer, M., Sanchez Paramo, C., January 2015 16 Managing risk with insurance and savings: experimental evidence for male and female farm managers in the Sahel Delavallade, C., Dizon, F., Hill, R., Petraud, J. P., el., January 2015 17 Gone with the storm: rainfall shocks and household well‐being in Guatemala Baez, J. E., Lucchetti, L., Genoni, M. E., Salazar, M., January 2015 18 Handling the weather: insurance, savings, and credit in West Africa De Nicola, F., February 2015 19 The distributional impact of fiscal policy in South Africa Inchauste Comboni, M. G., Lustig, N., Maboshe, M., Purfield, C., Woolard, I., March 2015 20 Interviewer effects in subjective survey questions: evidence from Timor‐Leste Himelein, K., March 2015 21 No condition is permanent: middle class in Nigeria in the last decade Corral Rodas, P. A., Molini, V., Oseni, G. O., March 2015 22 An evaluation of the 2014 subsidy reforms in Morocco and a simulation of further reforms Verme, P., El Massnaoui, K., March 2015 Updated on December 2015 by POV GP KL Team | 2 23 The quest for subsidy reforms in Libya Araar, A., Choueiri, N., Verme, P., March 2015 24 The (non‐) effect of violence on education: evidence from the "war on drugs" in Mexico Márquez‐Padilla, F., Pérez‐Arce, F., Rodriguez Castelan, C., April 2015 25 “Missing girls” in the south Caucasus countries: trends, possible causes, and policy options Das Gupta, M., April 2015 26 Measuring inequality from top to bottom Diaz Bazan, T. V., April 2015 27 Are we confusing poverty with preferences? Van Den Boom, B., Halsema, A., Molini, V., April 2015 28 Socioeconomic impact of the crisis in north Mali on displaced people (Available in French) Etang Ndip, A., Hoogeveen, J. G., Lendorfer, J., June 2015 29 Data deprivation: another deprivation to end Serajuddin, U., Uematsu, H., Wieser, C., Yoshida, N., Dabalen, A., April 2015 30 The local socioeconomic effects of gold mining: evidence from Ghana Chuhan-Pole, P., Dabalen, A., Kotsadam, A., Sanoh, A., Tolonen, A.K., April 2015 31 Inequality of outcomes and inequality of opportunity in Tanzania Belghith, N. B. H., Zeufack, A. G., May 2015 32 How unfair is the inequality of wage earnings in Russia? estimates from panel data Tiwari, S., Lara Ibarra, G., Narayan, A., June 2015 33 Fertility transition in Turkey—who is most at risk of deciding against child arrival? Greulich, A., Dasre, A., Inan, C., June 2015 34 The socioeconomic impacts of energy reform in Tunisia: a simulation approach Cuesta Leiva, J. A., El Lahga, A., Lara Ibarra, G., June 2015 35 Energy subsidies reform in Jordan: welfare implications of different scenarios Atamanov, A., Jellema, J. R., Serajuddin, U., June 2015 36 How costly are labor gender gaps? estimates for the Balkans and Turkey Cuberes, D., Teignier, M., June 2015 37 Subjective well‐being across the lifespan in Europe and Central Asia Bauer, J. M., Munoz Boudet, A. M., Levin, V., Nie, P., Sousa‐Poza, A., July 2015 Updated on December 2015 by POV GP KL Team | 3 38 Lower bounds on inequality of opportunity and measurement error Balcazar Salazar, C. F., July 2015 39 A decade of declining earnings inequality in the Russian Federation Posadas, J., Calvo, P. A., Lopez‐Calva, L.‐F., August 2015 40 Gender gap in pay in the Russian Federation: twenty years later, still a concern Atencio, A., Posadas, J., August 2015 41 Job opportunities along the rural‐urban gradation and female labor force participation in India Chatterjee, U., Rama, M. G., Murgai, R., September 2015 42 Multidimensional poverty in Ethiopia: changes in overlapping deprivations Yigezu, B., Ambel, A. A., Mehta, P. A., September 2015 43 Are public libraries improving quality of education? when the provision of public goods is not enough Rodriguez Lesmes, P. A., Valderrama Gonzalez, D., Trujillo, J. D., September 2015 44 Understanding poverty reduction in Sri Lanka: evidence from 2002 to 2012/13 Inchauste Comboni, M. G., Ceriani, L., Olivieri, S. D., October 2015 45 A global count of the extreme poor in 2012: data issues, methodology and initial results Ferreira, F.H.G., Chen, S., Dabalen, A. L., Dikhanov, Y. M., Hamadeh, N., Jolliffe, D. M., Narayan, A., Prydz, E. B., Revenga, A. L., Sangraula, P., Serajuddin, U., Yoshida, N., October 2015 46 Exploring the sources of downward bias in measuring inequality of opportunity Lara Ibarra, G., Martinez Cruz, A. L., October 2015 47 Women’s police stations and domestic violence: evidence from Brazil Perova, E., Reynolds, S., November 2015 48 From demographic dividend to demographic burden? regional trends of population aging in Russia Matytsin, M., Moorty, L. M., Richter, K., November 2015 49 Hub‐periphery development pattern and inclusive growth: case study of Guangdong province Luo, X., Zhu, N., December 2015 50 Unpacking the MPI: a decomposition approach of changes in multidimensional poverty headcounts Rodriguez Castelan, C., Trujillo, J. D., Pérez Pérez, J. E., Valderrama, D., December 2015 51 The poverty effects of market concentration Rodriguez Castelan, C., December 2015 52 Can a small social pension promote labor force participation? evidence from the Colombia Mayor program Pfutze, T., Rodriguez Castelan, C., December 2015 Updated on December 2015 by POV GP KL Team | 4 53 Why so gloomy? perceptions of economic mobility in Europe and Central Asia Davalos, M. E., Cancho, C. A., Sanchez, C., December 2015 54 Tenure security premium in informal housing markets: a spatial hedonic analysis Nakamura, S., December 2015 For the latest and sortable directory, available on the Poverty & Equity GP intranet site. http://POVERTY WWW.WORLDBANK.ORG/POVERTY Updated on December 2015 by POV GP KL Team | 5