Earnings Inequality and Risk over Two Decades of Economic Development in Lithuania
0 likes78 views
The document analyzes earnings inequality and dynamics in Lithuania over two decades, revealing a significant reduction in income inequality among workers aged 25-55 across both genders, particularly between income percentiles. It discusses the impact of Lithuania's economic development, including EU accession, on labor markets and income mobility. Key findings include the stability of income volatility compared to advanced economies and the unique income mobility patterns observed in the country.
![Dispersion of log earnings changes by permanent income
(a) Dispersion, Men
0
.5
1
1.5
2
P90-P10
Differential
of
g
it
0 10 20 30 40 50 60 70 80 90 100
Quantiles of Permanent Income Pit-1
[25-34]
[35-44]
[45-55]
(b) Dispersion, Women
0
.5
1
1.5
2
P90-P10
Differential
of
g
it
0 10 20 30 40 50 60 70 80 90 100
Quantiles of Permanent Income Pit-1
[25-34]
[35-44]
[45-55]
Note: H-sample, 1-year changes in residualized log earnings. Source: SoDra, 2000–2020.
17 / 33](https://image.slidesharecdn.com/presentation-250102081015-7fbc0f4b/85/Earnings-Inequality-and-Risk-over-Two-Decades-of-Economic-Development-in-Lithuania-18-320.jpg)
![Skewness of log earnings changes by permanent income
(a) Kelley skewness, Men
-.3
-.2
-.1
0
.1
.2
Kelley
Skewness
of
g
it
0 10 20 30 40 50 60 70 80 90 100
Quantiles of Permanent Income Pit-1
[25-34]
[35-44]
[45-55]
(b) Kelley skewness, Women
-.3
-.2
-.1
0
.1
.2
Kelley
Skewness
of
g
it
0 10 20 30 40 50 60 70 80 90 100
Quantiles of Permanent Income Pit-1
[25-34]
[35-44]
[45-55]
Note: H-sample, 1-year changes in residualized log earnings. Source: SoDra, 2000–2020.
18 / 33](https://image.slidesharecdn.com/presentation-250102081015-7fbc0f4b/85/Earnings-Inequality-and-Risk-over-Two-Decades-of-Economic-Development-in-Lithuania-19-320.jpg)
![Kurtosis of log earnings changes by permanent income
(a) Excess Crow-Siddiqui kurtosis, Men
1
3
5
7
9
11
13
Excess
Crow-Siddiqui
Kurtosis
of
g
it
0 10 20 30 40 50 60 70 80 90 100
Quantiles of Permanent Income Pit-1
[25-34]
[35-44]
[45-55]
(b) Excess Crow-Siddiqui kurtosis, Women
1
3
5
7
9
11
13
Excess
Crow-Siddiqui
Kurtosis
of
g
it
0 10 20 30 40 50 60 70 80 90 100
Quantiles of Permanent Income Pit-1
[25-34]
[35-44]
[45-55]
Note: H-sample, 1-year changes in residualized log earnings. Source: SoDra, 2000–2020.
19 / 33](https://image.slidesharecdn.com/presentation-250102081015-7fbc0f4b/85/Earnings-Inequality-and-Risk-over-Two-Decades-of-Economic-Development-in-Lithuania-20-320.jpg)
![10-year Income mobility over the life cycle
(a) Men
Top 0.1% of Pit-1
0
20
40
60
80
100
Mean
Percentiles
of
P
it+10
0 10 20 30 40 50 60 70 80 90 99
Percentiles of Permanent Income, Pit
[25-34]
[35-44]
(b) Women
Top 0.1% of Pit-1
0
20
40
60
80
100
Mean
Percentiles
of
P
it+10
0 10 20 30 40 50 60 70 80 90 99
Percentiles of Permanent Income, Pit
[25-34]
[35-44]
Note: H-sample, average rank-rank mobility for men and women of different ages. The black diagonal dashed line
is the 45-degree line corresponds to the case of no mobility. Source: SoDra, 2000–2020.
22 / 33](https://image.slidesharecdn.com/presentation-250102081015-7fbc0f4b/85/Earnings-Inequality-and-Risk-over-Two-Decades-of-Economic-Development-in-Lithuania-23-320.jpg)
![GDP betas by gender, age and permanent income
(a) Men
0
.5
1
1.5
2
GDP
Beta
0 10 20 30 40 50 60 70 80 90 100
Quantiles of Permanent Income Pit-1
[25-34]
[35-44]
[45-54]
(b) Women
0
.5
1
1.5
2
GDP
Beta
0 10 20 30 40 50 60 70 80 90 100
Quantiles of Permanent Income Pit-1
[25-34]
[35-44]
[45-54]
Note: H-sample, 1-year changes in log earnings. βg estimates. Shaded areas represent 95% confidence intervals
based on robust standard errors. Source: SoDra, 2000–2020.
30 / 33](https://image.slidesharecdn.com/presentation-250102081015-7fbc0f4b/85/Earnings-Inequality-and-Risk-over-Two-Decades-of-Economic-Development-in-Lithuania-31-320.jpg)
More Related Content
Similar to Earnings Inequality and Risk over Two Decades of Economic Development in Lithuania (20)
Earnings Inequality and Risk over Two Decades of Economic Development in Lithuania
- 1. Earnings Inequality and Risk over Two Decades of Economic Development in Lithuania Jose Garcia-Louzao1,2,3 Linas Tarasonis1,2 1Bank of Lithuania 2Vilnius University 3CESifo GRID 2.0 workshop November 8, 2024 The views expressed here do not necessarily reflect the position of the Bank of Lithuania or the Eurosystem
- 2. Outline Lithuania in GRID: Data and institutional context Insights from Lithuania: Common part Income Inequality Income Dynamics Income Mobility Insights from Lithuania: Country-specific analysis GDP betas Concluding remarks 1 / 33
- 3. Lithuania in GRID: Data - Administrative data from the State Social Insurance Fund Board (SoDra) - 25% random sample of Social Security population over 2000-2020 - based on month-year of birth → the sampling is not biased by migration - monthly records from 2010 onwards; quarterly records from 2000 to 2009 - workers: id, gender, age, insured labor income (base pay + non-regular payments, no top-coding), no hours - Estimation sample - convert data into an annual panel; income pooled across employers - drop self-employment (11%) and if earning less than 1.5 monthly minimum wage annually - individuals, aged 25-55: 0.68M workers, 7.3M worker-year observations 2 / 33
- 4. Institutional context in Lithuania between 2000 and 2020 - 2001: WTO Membership - Opened up the economy to international markets. - 2004: EU Membership - Brought significant political, economic, and social changes. - Access to EU funds for infrastructure, economic, and social policies. - Enhanced democracy and governance to meet EU standards. - Access to new trading partners and foreign investment. - Free Movement of Capital and Labor - Supported economic growth and labor mobility. - Led to a wave of emigration: over 5% of the working-age population moved to other EU countries by 2009 (Fic et al., 2011). - Emigration impacted the labor market and earnings dynamics in Lithuania. 3 / 33
- 5. Macro developments in Lithuania between 2000 and 2020 1 1.3 1.6 1.9 2.2 2.5 Index (2000=1) 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 Year GDP Productivity (a) Real GDP and labor productivity .2 .3 .4 .5 .6 .7 .8 Share of GDP 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 Year Exports Imports FDI (b) Openness Note: Real GDP and labor productivity (gross value added per worker) normalized to their value in 2000. Source: Statistics Lithuania and own calculations. 4 / 33
- 6. Labor developments in Lithuania between 2000 and 2020 1 1.6 2.2 2.8 3.4 4 Firms and employees (2000=1) 1 .97 .94 .91 .88 .85 Working-age population (2000=1) 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 Year Working-age population Firms (rhs) Employees (rhs) (a) Employment and firms .35 .38 .41 .44 .47 .5 Labor compensation over GDP 1 1.6 2.2 2.8 3.4 4 Index (2000=1) 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 Year Minimum wage Average wage Labor share (rhs) (b) Labor compensation Note: Working-age population, employment, firms, wages, and the minimum wage are normalized to their value in 2000. Source: Statistics Lithuania and own calculations. 5 / 33
- 7. Insights from Lithuania: Income Inequality 6 / 33
- 8. Percentiles of the distribution of log annual earnings (a) Overall, Men 0 .2 .4 .6 .8 1 1.2 Percentiles Relative to 2000 2000 2004 2008 2012 2016 2020 p90 p75 p50 p25 p10 (b) Overall, Women 0 .2 .4 .6 .8 1 1.2 Percentiles Relative to 2000 2000 2004 2008 2012 2016 2020 p90 p75 p50 p25 p10 Note: CS-sample. All percentiles are normalized to 0 in 2000. The shaded areas indicate recession years. Source: SoDra, 2000–2020. 7 / 33
- 9. Upper percentiles of the distribution of log annual earnings (a) Top percentiles, Men 0 .2 .4 .6 .8 1 Percentiles Relative to 2000 2000 2004 2008 2012 2016 2020 p99.9 p99 p95 p90 (b) Top percentiles, Women 0 .2 .4 .6 .8 1 Percentiles Relative to 2000 2000 2004 2008 2012 2016 2020 p99.9 p99 p95 p90 Note: CS-sample. All percentiles are normalized to 0 in 2000. The shaded areas indicate recession years. Source: SoDra, 2000–2020. 8 / 33
- 10. Earnings inequality (a) Overall, Men 1.8 1.9 2 2.1 2.2 2.3 2.4 2.5 2.6 Dispersion of Log Earnings 2000 2004 2008 2012 2016 2020 2.56*σ P90-P10 (b) Overall, Women 1.8 1.9 2 2.1 2.2 2.3 2.4 2.5 2.6 Dispersion of Log Earnings 2000 2004 2008 2012 2016 2020 2.56*σ P90-P10 Note: CS-sample, σ denotes the standard deviation of log real annual earnings. The shaded areas indicate recession years. Source: SoDra, 2000–2020. 9 / 33
- 11. Earnings inequality: Right and left tails (a) Right and left tails, Men .6 .8 1 1.2 1.4 1.6 Dispersion of Log Earnings 2000 2004 2008 2012 2016 2020 P90-P50 P50-P10 (b) Right and left tails, Women .6 .8 1 1.2 1.4 1.6 Dispersion of Log Earnings 2000 2004 2008 2012 2016 2020 P90-P50 P50-P10 Note: CS-sample, σ denotes the standard deviation of log real annual earnings. The shaded areas indicate recession years. Source: SoDra, 2000–2020. 10 / 33
- 12. Life-cycle earnings inequality (a) Earnings profiles by cohort, Men 25 yrs old 35 yrs old 1.7 1.9 2.1 2.3 2.5 2.7 P90-P10 of Log Earnigs 2000 2004 2008 2012 2016 2020 Cohort 2000 Cohort 2006 Cohort 2010 Cohort 2016 (b) Earnings profiles by cohort, Women 25 yrs old 35 yrs old 1.7 1.9 2.1 2.3 2.5 2.7 P90-P10 of Log Earnigs 2000 2004 2008 2012 2016 2020 Cohort 2000 Cohort 2006 Cohort 2010 Cohort 2016 Note: CS-sample, Panels A and B consider only workers aged 25. Panels C and B plot earnings profiles by cohorts and age groups. Source: SoDra, 2000–2020. 11 / 33
- 13. Earnings inequality: GRID 1.0 context -.2 -.1 0 .1 .2 Dispersion of (log) annual earnings, P90-P10 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 Year ARG BRA CAN DKK ESP FRA GER ITA NOR SWE USA LTU 12 / 33
- 14. Income Inequality: Takeaways - Income inequality among workers aged 25-55 has decreased significantly: - Across both genders - For both upper and lower tails of the income distribution - Within and between cohorts - This reduction in income inequality is nearly unprecedented, especially among GRID 1.0 countries. 13 / 33
- 15. Insights from Lithuania: Income Dynamics 14 / 33
- 16. Dispersion of 1-year log earnings changes (a) Upper and lower dispersion, Men .3 .4 .5 .6 .7 .8 .9 1 1.1 1.2 Dispersion of g 1 it 2000 2004 2008 2012 2016 2020 P90-P50 P50-P10 (b) Upper and lower dispersion, Women .3 .4 .5 .6 .7 .8 .9 1 1.1 1.2 Dispersion of g 1 it 2000 2004 2008 2012 2016 2020 P90-P50 P50-P10 Note: LS-sample, 1-year changes in residualized log earnings. The shaded areas indicate recession years. Source: SoDra, 2000–2020. 15 / 33
- 17. Skewness and kurtosis of log earnings changes (a) Kelley Skewness -.5 -.4 -.3 -.2 -.1 0 .1 .2 Skewness of g 1 it 2000 2004 2008 2012 2016 2020 Women Men (b) Excess Crow-Siddiqui kurtosis 2 3 4 5 6 7 8 9 10 Excess Kurtosis of g 1 it 2000 2004 2008 2012 2016 2020 Women Men Note: LS-sample, 1-year changes in residualized log earnings. The shaded areas indicate recession years. Source: SoDra, 2000–2020. 16 / 33
- 18. Dispersion of log earnings changes by permanent income (a) Dispersion, Men 0 .5 1 1.5 2 P90-P10 Differential of g it 0 10 20 30 40 50 60 70 80 90 100 Quantiles of Permanent Income Pit-1 [25-34] [35-44] [45-55] (b) Dispersion, Women 0 .5 1 1.5 2 P90-P10 Differential of g it 0 10 20 30 40 50 60 70 80 90 100 Quantiles of Permanent Income Pit-1 [25-34] [35-44] [45-55] Note: H-sample, 1-year changes in residualized log earnings. Source: SoDra, 2000–2020. 17 / 33
- 19. Skewness of log earnings changes by permanent income (a) Kelley skewness, Men -.3 -.2 -.1 0 .1 .2 Kelley Skewness of g it 0 10 20 30 40 50 60 70 80 90 100 Quantiles of Permanent Income Pit-1 [25-34] [35-44] [45-55] (b) Kelley skewness, Women -.3 -.2 -.1 0 .1 .2 Kelley Skewness of g it 0 10 20 30 40 50 60 70 80 90 100 Quantiles of Permanent Income Pit-1 [25-34] [35-44] [45-55] Note: H-sample, 1-year changes in residualized log earnings. Source: SoDra, 2000–2020. 18 / 33
- 20. Kurtosis of log earnings changes by permanent income (a) Excess Crow-Siddiqui kurtosis, Men 1 3 5 7 9 11 13 Excess Crow-Siddiqui Kurtosis of g it 0 10 20 30 40 50 60 70 80 90 100 Quantiles of Permanent Income Pit-1 [25-34] [35-44] [45-55] (b) Excess Crow-Siddiqui kurtosis, Women 1 3 5 7 9 11 13 Excess Crow-Siddiqui Kurtosis of g it 0 10 20 30 40 50 60 70 80 90 100 Quantiles of Permanent Income Pit-1 [25-34] [35-44] [45-55] Note: H-sample, 1-year changes in residualized log earnings. Source: SoDra, 2000–2020. 19 / 33
- 21. Income Dynamics: Takeaways - Income volatility - Stable over time and higher than in advanced European economies (e.g., Germany, Sweden), comparable to the US and Brazil - Lower for higher-income groups, except for top earners - Higher for young workers, especially young female workers - Positively skewed for low-income workers and negatively skewed for high-income workers - Stable (resp., divergent) excess kurtosis between the 20th and 90th percentiles of permanent income for males (resp., females), unique to Lithuania - Cyclical pattern consistent with GRID 1.0 countries. 20 / 33
- 22. Insights from Lithuania: Income Mobility 21 / 33
- 23. 10-year Income mobility over the life cycle (a) Men Top 0.1% of Pit-1 0 20 40 60 80 100 Mean Percentiles of P it+10 0 10 20 30 40 50 60 70 80 90 99 Percentiles of Permanent Income, Pit [25-34] [35-44] (b) Women Top 0.1% of Pit-1 0 20 40 60 80 100 Mean Percentiles of P it+10 0 10 20 30 40 50 60 70 80 90 99 Percentiles of Permanent Income, Pit [25-34] [35-44] Note: H-sample, average rank-rank mobility for men and women of different ages. The black diagonal dashed line is the 45-degree line corresponds to the case of no mobility. Source: SoDra, 2000–2020. 22 / 33
- 24. 10-year Income mobility over time (a) Men Top 0.1% of Pit-1 0 20 40 60 80 100 Mean Percentiles of P it+10 0 10 20 30 40 50 60 70 80 90 99 Percentiles of Permanent Income, Pit 2005 2010 (b) Women Top 0.1% of Pit-1 0 20 40 60 80 100 Mean Percentiles of P it+10 0 10 20 30 40 50 60 70 80 90 99 Percentiles of Permanent Income, Pit 2005 2010 Note: H-sample, average rank-rank mobility for men and women, using two alternative base years 2005 and 2010 and averaging over all age groups. The black diagonal dashed line is the 45-degree line corresponds to the case of no mobility. Source: SoDra, 2000–2020. 23 / 33
- 25. Key Income mobility statistics A. Pooled, 2002-2021 RRS AUM ADM M99 0.68 33.1 66.9 93.6 B. AUM 2002 2007 2010 2015 33.7 33.0 32.9 33.0 C. AUM Pooled, 2002-2021 All 25-34 35-44 45-55 33.1 36.0 32.5 31.0 D. AUM Pooled, 2002-2020, Men All 25-34 35-44 45-55 33.1 34.1 32.8 32.3 E. AUM Pooled, 2002-2020, Women All 25-34 35-44 45-55 33.2 37.8 32.1 29.9 Note: LS-sample, 5-year changes in log earnings. RSS: rank-rank slope; AUM: absolute upward mobility, i.e. expected rank at t + 5 conditional on being below the median at time t; ADM: absolute downward mobility, i.e. expected rank at t + 5 conditional on being above the median at time t; M99: expected rank at t + 5 conditional on being in the top 1% at time t. Source: SoDra, 2000–2020 24 / 33
- 26. Income mobility: Great Gatsby curve ARG BRA CAN DKK ESP FRA GER ITA MEX NOR SWE USA UK LTU 2002-2007 LTU 2015-2020 .65 .7 .75 .8 .85 .9 Rank-Rank Slope .25 .3 .35 .4 .45 .5 .55 .6 Gini Coefficient Source: SoDra, 2002-2020 for Lithuania, GRID 1.0 results for other countries. 25 / 33
- 27. Income Mobility: Takeaways - Income mobility in Lithuania is characterized by: - High levels overall - Stability over time - Significantly higher mobility among young individuals - High mobility among young females, contrasted with low mobility among older females - Lithuania’s combination of high income mobility and high income inequality positions it as an outlier on the Great Gatsby Curve among GRID 1.0 countries. 26 / 33
- 28. Insights from Lithuania: GDP betas 27 / 33
- 29. GDP Betas - We relate individual labor earnings changes to GDP growth following Guvenen et al. (2017): ∆yit = αg + βg∆Yt + ϵit (1) where - yit denotes the (log) real annual earnings of individual i in year t, - Yt is the (log) real GDP in year t, - βg is the GDP beta of workers in group g. - We also follow Busch et al. (2022) to correlate distributional moments of ∆yit with the GDP growth. - Groups g defined by: - sex, - three age categories, - 20 percentiles of the permanent labor earnings distribution. 28 / 33
- 30. GDP betas by gender (1) (2) (3) (4) Growth P90-P10 Kelley Kurtosis A. Men GDP beta 1.300*** -1.046*** 1.816*** 9.895*** (0.006) (0.147) (0.120) (2.299) B. Women GDP beta 0.669*** -0.135 0.997*** 5.613*** (0.004) (0.106) (0.076) (1.398) Note: H-sample, 1-year changes in log earnings. The g-groups are defined by sex only. Robust standard errors in parentheses. Source: SoDra, 2000–2020 29 / 33
- 31. GDP betas by gender, age and permanent income (a) Men 0 .5 1 1.5 2 GDP Beta 0 10 20 30 40 50 60 70 80 90 100 Quantiles of Permanent Income Pit-1 [25-34] [35-44] [45-54] (b) Women 0 .5 1 1.5 2 GDP Beta 0 10 20 30 40 50 60 70 80 90 100 Quantiles of Permanent Income Pit-1 [25-34] [35-44] [45-54] Note: H-sample, 1-year changes in log earnings. βg estimates. Shaded areas represent 95% confidence intervals based on robust standard errors. Source: SoDra, 2000–2020. 30 / 33
- 32. GDP betas: Takeaways - GDP betas in Lithuania are characterized by: - Generally high levels of sensitivity to GDP fluctuations - Greater sensitivity for men compared to women - Strong negative correlation between GDP growth and the volatility of earnings changes for men, but not for women - No U-shape pattern: Sensitivity decreases with higher permanent income levels, contrasting with trends observed in the US. 31 / 33
- 33. Insights from Lithuania: Concluding remarks 32 / 33
- 34. Concluding remarks Overall, Lithuanian economy exhibits the following key characteristics: - Significant reduction in income inequality, reaching nearly unprecedented levels, especially among GRID 1.0 countries. - High income volatility, remaining stable over time with cyclical patterns consistent with GRID 1.0 countries. - Unique combination of high income mobility and high income inequality, coexisting within the economy. - Absence of a U-shape pattern in GDP betas across permanent income levels, distinguishing Lithuania from other countries. 33 / 33