Uruguay: Selected Issues

Abstract

Uruguay: Selected Issues

Increased Social Inclusion—the Role of Government Policies1

This chapter analyzes the lowering of the Gini coefficient in Uruguay during the last six years. Social policies and transfers have played a significant role in reducing poverty and inequality. While income dispersion has decreased across Latin America over the last decade, Uruguay stands out as the country with the largest drop in the Gini coefficient between 2009 and 2014, and to the lowest level. This reflects both government guidelines to bolster low wages, and increased redistribution through income taxes and transfers. However, looking ahead, the positive effects of further redistributive policies may be weighed against their fiscal costs and by a possible trade-off between income compression and incentives for labor supply and education and training. Work incentives among women can be strengthened further via reforms of parental leave, to reduce the remaining gender wage gap, and would diminish future pressures on public finances due to population ageing.

A. Introduction

1. According to recent IMF studies2, lower income net inequality is robustly correlated with faster and more durable economic growth. Up to a certain point3, it increases the potential growth rate which is especially important given the long term challenges of an ageing population4. The main arguments behind these positive effects are that inequality can undermine progress in health and education, and cause investment-reducing political instability and economic instability. It can also undercut the social consensus required to adjust in the face of shocks. A higher potential output would also imply less pressure on public finances in the future, by increasing public income and dampening public expenditures.

2. There has been a downward trend in the Gini coefficient across countries in Latin America during the 2000s. This decline was to a large extent due to high economic growth that lifted low-income earners wages5, in many cases due to increased demand for low skilled workers during the commodity price boom6. For Uruguay as well, our calculations confirm that inequality has dropped since 2007, and especially since 2009 (see text chart)—based on microdata from the annual household survey. An important question is then: what are the main forces behind the decrease in the Gini coefficient in Uruguay? How much have government policies contributed?

A04ufig1

Household Net Per Capita Gini

(From 0 to 1)

Citation: IMF Staff Country Reports 2017, 029; 10.5089/9781475573626.002.A004

Source: SEDLAC/World Bank, Instituto Nacional de Estadistica Uruguay (INE) and Fund staff calculations using data from yearly houshold survey data from “Encuesta Continua De Hogares, INE.1)The difference between the level of Gini coefficient between the calculations done by SEDLAC/World Bank/IMF and INE (of Uruguay is mainly explained by adjustments by including the FONASA health care system in the latter example. This incorporation lowers the Gini coefficient.2) The average of LAC6 countries (Argentina, Brazil, Chile, Colombia, Mexico and Peru) includes data inter- and extrapollations due to data gaps.

3. Social welfare policies have long played an important role in Uruguay, and government involvement has intensified during the last six years. Uruguay has been described as South America’s “first welfare state” as a result of its pioneering efforts in public education, health care, and social security. During the last six years the government has aimed to lower poverty and income inequalities further. The triparty wage negotiation policy7 has, among other things, focused on increasing wages among low income earners, and there has also been an increase in government transfers to households as well as in income taxes for high-income earners.

Two Approaches to Calculating Contributions to Changes in Income Inequality

This paper applies two alternative methods for determining the magnitudes of the factors behind the overall change in the Gini coefficient between two years:

A.) One fairly simple model is to start with the Gini coefficient based on gross market income (the sum of gross labor and gross capital income), and then to recalculate the Gini coefficient after incorporating, one by one, income taxes, social contributions, public transfers etc. For each consecutive step, the difference between these—before and after—coefficients can be viewed as the contribution from the additional variable for a specific year. The contribution effect can also be compared between years in order to evaluate changes in the redistributional effect. However, with this simple approach, the results can be dependent on the sequencing of the added variables.1

B.) An alternative approach is to apply a number of different paths for sequencing the additional variable, and take the average result. In the calculations using this approach in this paper, the average of 720 paths are used. This approach only calculates the contribution from different income sources and taxes between two years.

In this study, both methods are used, as the first one can be easier to follow and evaluate, and the second is more robust.

It has to be underlined that these results should be interpreted with some caution. They only include static effects, not dynamic effects, for instance that the need for public transfers might be smaller when economic growth is high, or that different instruments can have different impacts on labor market functioning.

1 See Essama-Nssah (2012), Fortin et al (2011) and Ferreira (2010).

4. More compressed labor income was the main factor behind the lower Gini coefficient in 2015 compared with 2009. The decline in the Gini coefficient amounted to about 5 percentage points for both the overall household net per capita income Gini and the “market per capita income” Gini, which indicates that pre-tax gross income played a significant role. This is also confirmed by the results of the calculations of the contributions to the lower Gini coefficient using the robust approach model B (see Box 1; and see text chart). According to this model, more compressed gross labor income actually contributed more (120 percent) than the measured change in the net per capita Gini coefficient. Accordingly, a clearly lower dispersion in gross market income appears to have been the main driving factor. 8 Incomes policy seems to have played a major role in this. The government determines a national minimum wage, which sets a floor for the tripartite wage negotiation process. This national minimum wage increased by 40 percent in real terms during 2009–2015. In addition, the wage guidelines prepared by the authorities for the current wage round proposes a faster-than-average increase for wages at the lower end of the distribution. In this context, that data also show that hourly incomes for low-skilled labor9 increased by almost 35 percent in real terms during 2009–2015, for medium-skilled labor by around 20 percent, but were more or less unchanged for high-skilled labor. The proportion of women and men earning less than the minimum wage decreased between 2009 and 2015 (from around 8 percent to around 5 percent). Interestingly, the number of low skilled workers appears to have decreased significantly between 2009 and 2015 with increased high skilled labor10.

A04ufig2

Contributionsto Lower Household Net Per Capita Gini between 2009 and 2015

(Percentage of total change, negative numbersare decreasing effects)

Citation: IMF Staff Country Reports 2017, 029; 10.5089/9781475573626.002.A004

Source: Fund staff calculations using adecomp. adecomp implements the shapley decomposition of changes in a welfare indicator as proposed by Azevedo, Sanfelicce and Minh (2012). Following Barros et al (2DD6), this method takes advantage of the additivity property of a welfare aggregate to construct a counterfactual unconditional distribution of the welfare aggregate by changing each component at a time to calculate their contribution to the observed changes in poverty and inequality. Azevedo, Joao Pedro Viviane Sanfelice and Minh Cong Nguyen (2012) Shapley Decomposition by Components of a Welfare Measure. World Bank, (mimeo).Data from Instituto Nacionalde Estadistica Uruguay (INE).

5. Redistribution through taxes and transfers on balance has had a minor effect in promoting equality. According to the simple approach (model A, see Box 1.) labor-income taxes decreased the Gini coefficient by a little more than 1.5 percentage points in 2015 (see text chart)– a larger impact than in 2009. In particular, the inclusion of a seventh labor income tax bracket at 30 percent in 2013 (previously, the highest was a sixth bracket at 25 percent) for high income earners has contributed to lower net income dispersion. The contribution to the lower Gini coefficient between 2009 and 2015 from higher labor-income taxes is also confirmed in the robust model (see the text chart with paragraph 4). Payments for social contributions, on the other hand, contributed to widen the Gini coefficient, especially in 2015 (again, confirmed by the earlier chart for the robust model). In particular, there is an income-based cap on social contributions, and the share of high-skilled labor (with higher incomes) increased between these years. Government transfers11 also contributed to lower the Gini coefficient. They increased by almost 30 percent in real terms between the years examined, which contributed to lift income among the most vulnerable in society, although the effect on the Gini coefficient was fairly small.

A04ufig3

Redistributional Effects On Household Net Per Capita Gini

(Percentage points, negative numbers are decreasing effects)

Citation: IMF Staff Country Reports 2017, 029; 10.5089/9781475573626.002.A004

Source: Fund staff calculations using data from the yearly household survey data base “Encuesta Continua De Hogares. INE.Note: Public transfers include unemployment benefits, benefits for sick leave maternity leave, housing allowances, debit cards from the family allowance program (AFAM) and pensions.

6. As in other Latin American countries there has for decades been an increase in female labor force participation in Uruguay. It increased from around 45 percent in 1990 to about 55 percent today. During the economic slowdown since 2014, female employment and participation rates remained stable, while these rates fell for men. The share of highly-educated women in the labor force is particularly high.

A04ufig4

Labor Force Participation Ratio

(Percent of total)

Citation: IMF Staff Country Reports 2017, 029; 10.5089/9781475573626.002.A004

Sources: INE and Fund staff calculations.

7. Similar to other countries, there still exist a gender wage gap, however. Men are better paid than women at almost all education levels, although the gap has declined since 2009. The gender gap as the ratio of female to male wages, is low for the very lowest skilled workers with an education of up to 5 years, but increases thereafter at higher education levels. For low and medium skill levels, men earn about 20 percent more per hour. However, in 2015 the per hour gender wage gap had decreased significantly for high-skilled labor and was actually close to zero12. In particular, per hour wages for women in 2015 were close to those for men for those with 14 to 21 years of education (see text chart). Beyond that, the gap increases again, but the number of people with education of 22 years or above is small and contributes little to overall wage levels. The contribution of women to Uruguay’s overall household net per-capita Gini coefficient is slightly negative, that is, it decreases the Gini coefficient13. This decreasing effect was around 0.4 percentage points in both years.

A04ufig5

Male and Female Wages

(In pesos and years of education)

Citation: IMF Staff Country Reports 2017, 029; 10.5089/9781475573626.002.A004

Sources: Fund staff calculations using data from the yearly household survey data base “Encuesta Continua De Hogares, INE.

B. Policies to Support Inclusion

8. Policies to foster inclusive growth should remain an important part of Uruguay’s policy agenda. Redistribution can be a helpful tool to reduce inequality and poverty. However, it is not a substitute for policies to bolster directly generalized skills formation, gender equality, and labor market efficiency. Indeed, the finding that recent changes in income distribution were driven by workers’ primary sources of income underscores this message.

9. There are no definite signs so far of negative incentive effects on labor supply of the government’s policies that have compressed the distribution of wages. Both the employment ratio and the labor force participation rates increased between 2009 and 2015. The employment ratio increased by half of a percentage point to 59 percent and labor force participation by 0.7 percentage points to 63.8 percent. On the other hand, staff calculations based on the yearly household surveys, show a fairly large drop between 2009 and 2015 in people’s willingness to add labor hours to their current hours. The percentage of men 15–63 years of age who wanted to add working hours dropped from 15 percent to 10 percent and from 12 percent to 9 percent for women. Also, the actual average work week fell from 39.5 hours to 38.5 hours, which could be a sign of some adverse impact on work incentives.

10. Falling skill premiums could pose a risk for the willingness to develop human capital. Although dropout rates in secondary education have increased, the average number of years of education has continued to increase lately, especially among women (see text chart). Women surpassed men in terms of average years of education already in the early 1990s and this trend has continued since. However, there has been a significant decline in the income premium for more educated workers since 2010 (see text chart), especially for men, and most strongly for men with a high (compared with medium) level of education. This decline in the skills premium could be the result of the increased availability of educated workers14, but over time it could further reduce the incentives for men to improve their skills.

A04ufig6

Average Education

(Years)

Citation: IMF Staff Country Reports 2017, 029; 10.5089/9781475573626.002.A004

Sources: Socio-Economic Database for Latin America and the Caribbean (CED LAS and The World Bank).
A04ufig7

Skill premium

Citation: IMF Staff Country Reports 2017, 029; 10.5089/9781475573626.002.A004

Sources: Fund staff calculations using data from the yearly household survey data base “Encuesta Continua De Hogares, INE.

11. Policies should also focus on improving efficiency in secondary education. Uruguay has a higher teacher-to-student ratio than other comparable Latin American countries, but just marginally higher PISA-studies results (see text chart), following some improvement in 2015. A further challenge is the improvement of the coverage and quality of vocational training15.

A04ufig8

Teachers and Outcome, Secondary Education, Latest Value Available

Citation: IMF Staff Country Reports 2017, 029; 10.5089/9781475573626.002.A004

Source: IMF FAD Expenditure Assessment Tool (EAT); OECD, Pisa 2015 database and World Bank WDI database.

12. Policies to further increase female labor force participation could limit the economic challenges of an ageing population. Boosting female labor participation would help offset the future drop in labor supply due to population ageing16 and address the need to further increase skills-intensive labor supply17. Currently, the participation gap is still as high as 17 percentage points. In particular, the proportion of young women neither employed nor studying is high18. Policies are already in place to encourage women to work, including affordable child care for low-income households. Although the percentage of families that use child-care facilities is fairly low compared to OECD countries, it is far higher than many other countries in Latin America. Furthermore, the government has proposed to further enhance affordable child care in the coming years. As an additional step, parental leave could be enhanced further. In particular, the length of maternity leave could be increased from todays’ twelve weeks, paid paternity leave could be extended, and “daddy months” could be introduced to facilitate womens’ efforts to combine their careers with taking care of their children in a more balanced way. It could also decrease the gender wage gap19, which would further strengthen womens’ incentives to enter the labor market.

C. Conclusions

13. Uruguay has a long history of social welfare policies, and reforms have sought to raise people out of poverty and to lower inequality. Wage policies, increased transfers, and the labor-income tax for high wage earners have contributed to this. Greater social inclusion can increase the potential for economic growth. However, looking forward, further measures to increase social conclusion have to be assessed and weighed against their potential impact on incentives to develop skills and seek work. In addition, education reform could enhance the system’s efficiency. In light of the long-term challenges of an ageing population, female labor force participation could be further encouraged.

References

  • Azevedo, J.P., Gabriela Inchaust and Viviane Sanfelice, Policy Research Working Paper 6715, “Decomposing the Recent Inequality Decline in Latin America”, The World Bank, December 2013.

    • Search Google Scholar
    • Export Citation
  • Essama-Nssah, B. (2012). “Identification of Sources of Variation in Poverty Outcomes”, World Bank Policy Research Working Papers, No. 5954.

    • Search Google Scholar
    • Export Citation
  • Ferreira, Francisco H.G. (2010) “Distributions in Motion: Economic Growth, Inequality and Poverty Dynamics”. World Bank Policy Research Working Paper No. 5424. The World Bank, Washington, D.C.

    • Search Google Scholar
    • Export Citation
  • Fortin Nicole, Lemieux Thomas and Firpo Sergio. (2011). “Decomposition Methods in Economics”. In: Ashenfelter Orley and Card David (eds) Handbook of Labor Economics, Vol. 4A, pp. 1102. Northolland, Amsterdam.

    • Search Google Scholar
    • Export Citation
  • Gasparini, L., Cruses, G. and Tornarolli, L. (2011). “Recent Trends in Income Inequality in Latin America”. Economia 10, 147201.

  • IMF Staff Discussion Note (2014), “Redistribution, Inequality and Growth”, February 2014.

  • IMF Staff Discussion Note (2015), “Causes and Consequences of Income Inequality: A Global Perspective”, June 2015.

  • IMF Staff Discussion Note (2015), “Catalyst for Change: Empowering Women and Tackling Income Inequality”, October 2015.

  • IMF Working PaperInequality and Growth: A Heterogeneous Approach”, Grigoli, F., Evelio Paredes and Gabriel Di Bella (2016).

  • Johansson, Elly-Ann,The Effect of Own and Spousal Parental Leave on Earnings” (2010), Institute of Evaluation of Labour Market and Education Policy, Sweden, Working Paper 2010:4.

    • Search Google Scholar
    • Export Citation
  • Ministry of Labor and Social Security (2014), “Fifth Round of Wage Negotiations – 2013-2014”, April 2014, Labor Relations and Employment Evaluation and Monitoring Unit.

    • Search Google Scholar
    • Export Citation
  • OECD (2014), “Deveopment Pathways “Multi-dimentional Review of Uruguay”, Volume 1, Initial Assessment, 2014.

1

Prepared by Bengt Petersson.

2

“Redistribution, Inequality and Growth”, IMF Staff Discussion Note, February 2014, and “Causes and Consequences of Income Inequality: A Global Perspective”, IMF Staff Discussion Note, June 2015.

3

Grigoli et al. (2016) show that the positive effects on growth from lowering the Gini coefficient are significant in the Middle East and Central Asia, the Western Hemisphere, and across emerging markets in general (regions characterized by high initial income inequality).

5

See Azevedo, Inchaust and Sanfelice, (2013).

7

The tripartite wage setting is a process between employers, trade unions and the government, and guided by official government guidelines.

8

According to Ministry of Labor and Social Security (2014), the real adjustment for each year in 2012-2015 was higher the lower the income level was, a trend that was already observed in 2010-2011. Moreover, in the latest wage negotiation round, increases in wages are again clearly higher for the lowest income earners than for those with higher incomes.

9

Low-skilled labor is here defined as less than 9 years of education, medium-skilled as between 9 and 13 years and high-skilled as 14 years or longer. Average per hour labor income was in 2015 slightly below US$4.

10

According to our calculations, using the yearly household surveys for these two years, the number of low-skilled workers decreased by more than 20 percent, but increased by almost as much for high-skilled labor. However, we found no evidence explaining to what extent this transformation is due to higher demand for high-skilled labor versus increased supply.

11

These include unemployment benefits, benefits for sick leave, maternity leave, housing allowances, debit cards from the family allowance program (AFAM), and pensions.

12

The per hour wage gap for highly skilled labor decreased from 13 percent in 2009 to only 1 percent in 2015.

13

These calculations are simply to compare the Gini coefficients with and without women. See Gonzales, Jain-Chandra, Kochhar, Newiah, and Zeinullayev (2015) for a broader analysis of the linkages between gender and overall income inequality.

14

The number of male labor with high education increased by around 20 percent. For women the increase was around 15 percent.

16

Uruguay has a rapidly ageing population with shares similar to European and North American countries, see OECD (2014).

18

61 percent according to OECD (2014).

19

See Johansson, Elly-Ann, “The Effect of Own and Spousal Parental Leave on Earnings” (2010), Institute of Evaluation of Labour Market and Education Policy, Sweden, Working Paper 2010:4.

Uruguay: Selected Issues
Author: International Monetary Fund. Western Hemisphere Dept.
  • View in gallery

    Household Net Per Capita Gini

    (From 0 to 1)

  • View in gallery

    Contributionsto Lower Household Net Per Capita Gini between 2009 and 2015

    (Percentage of total change, negative numbersare decreasing effects)

  • View in gallery

    Redistributional Effects On Household Net Per Capita Gini

    (Percentage points, negative numbers are decreasing effects)

  • View in gallery

    Labor Force Participation Ratio

    (Percent of total)

  • View in gallery

    Male and Female Wages

    (In pesos and years of education)

  • View in gallery

    Average Education

    (Years)

  • View in gallery

    Skill premium

  • View in gallery

    Teachers and Outcome, Secondary Education, Latest Value Available