Technical Note
The Quality of Jobs
in Latin America and the Caribbean
Karen Barreto, Hernan Winkler, Carolina Diaz Bonilla, Diana Sanchez
October 2024
Summary
The creation of more and better jobs has been a key driver of poverty and inequality reduction.
However, while estimates on the number of jobs are available for most countries,
comprehensive measures of job quality are not typically reported systematically. This note
contributes to the filling of this gap by studying patterns of job quality across countries and over
time in Latin America and the Caribbean (LAC). This study uses the Job Quality Index (JQI) based
on Brummundi, Mann, and Rodriguez-Castelan (2018), which incorporates four key
employment characteristics that are key to assessing job quality: earnings, benefits, security,
and satisfaction.
Key messages
• The JQI exhibits substantial variation across countries, with Chile and Costa Rica leading
the region.
• Important gaps in job quality also exist within countries, with women, youth, and rural
workers having lower-quality jobs.
• The employment-to-GDP elasticity grows in most countries when adjusting for job
quality; that is, economic growth is not only associated with job creation, but also with
the creation of better jobs.
• Episodes of poverty and inequality reduction tend to be accompanied with increases in
job quality.
• Low labor productivity is linked to poorer job quality across the region.
1
1. Motivation and Methods
Changes in employment and earnings have been one of the main drivers of poverty and inequality
reduction in LAC over the last decade.1 However, the role of jobs in reducing poverty and inequality
goes beyond the level of earnings associated with them. Having social insurance coverage associated
with employment can be an important tool for preventing the vulnerable from falling into poverty as
they grow old or if they become sick. Accordingly, having stable employment helps protect earnings
from the ups and downs of the business cycle. Finally, having a job that is empowering and rewarding
can be welfare enhancing on its own beyond the associated monetary compensation (World Bank 2013).
The task of assigning a monetary value to these job characteristics involves a high degree of
ambiguity. While the market value of pension, disability, and health benefits or a more secure job can
be large, the calculation of each involves several complex decisions. For instance, it requires making
strong assumptions about the timing, likelihood, and welfare impacts of illness and family-care
responsibilities and how different types of jobs may mitigate or exacerbate to varying degrees such
direct impacts. At the same time, the likelihood, timing, and size of these shocks vary across age, area of
residence (urban/rural) and educational attainment, as well as gender. This challenge becomes even
harder when trying to estimate job quality measures that are comparable across countries and across
time.
This article adopts the approach to the analysis of job quality developed by Brummundi, Mann, and
Rodriguez-Castelan (2018), who aggregate different dimensions of a job into one index, based on a
methodology similar to that of multidimensional poverty measurement. This study includes four
dimensions to measure job quality. First, it considers whether a job pays high enough earnings, that is,
above a minimum level of well-being. Second, it considers whether a job provides benefits such as
health insurance and old-age pensions.2 Third, it considers job security, which aims to capture the extent
to which workers are protected from becoming poorer during economic downturns. In addition, job
security is associated with other positive outcomes such as reduced stress and improved mental health
(Watson and Osberg 2018; LaMontagne et al. 2020). Finally, it considers job satisfaction in order to
capture the role of other nonmonetary or unobservable characteristics of job quality. For example, our
data sets do not allow for the systematic measuring of on-the-job training or expectations about career
growth, which can be important factors for evaluating job quality.
In practice, these four dimensions of job quality are measured according to the following criteria:
1. Labor Income (������������������������������������������������������ ): labor market earnings above the poverty line of US$6.85/day at 2017
Purchasing Power Parity (PPP), which is the upper-middle-income-country poverty line.
2. Benefits (������������������������������������������������������������������ ): the job provides health insurance or retirement benefits.
3. Security (������������������������������������������������������������������ ): the job is considered secure if either (1) the worker has a contract, (2) the
job is permanent, or (3) the worker has kept the job for a long enough period (at least three
years).
4. Satisfaction (������������������������������������������������������������������������������������������ ): Given that not all surveys contain questions on job satisfaction, we
presume that the worker is satisfied with their job if they do not have a second job. This is based
on literature that has documented the association between the holding of one or more
1 Source: LAC Equity Lab, accessed October 2, 2024.
2 While several countries in Latin America already provide a minimum level of universal health-care and old-age pension coverage,
these benefits may be of lower quality than those associated with jobs.
2
additional jobs as proxy variables for low levels of job satisfaction in the main job, because they
are typically associated with earnings and hours-of-work constraints, insecurity, volatility, and
precarious work conditions, as well as increased physical hardship and poor mental health
outcomes (Dickey, Watson, and Zangelidis 2010; Zangelidis 2014; Bruns and Pilkauskas 2019;
Klinger and Weber 2020; Pouliakas and Conen 2023).
The JQI is then constructed according to the Alkire and Foster (2011) framework for creating a
multidimensional index. It requires that each indicator for every observation be treated as either a
success or a failure. Failures are given a 0, while successes are given a 1. It is important to mention that
all the observations used are from individuals who are in the labor force. That is, we include employed
workers and unemployed workers (those who are actively seeking work). If a worker is unemployed, all
dimensions are equal to 0 and so is the JQI. Accordingly, if labor earnings are below the poverty line, the
JQI is also equal to 0. In other words, a minimum level of earnings is a necessary condition for having a
high-quality job. For example, the methodology treats a job with earnings below the poverty line as a
low-quality job even if the worker has a secure job with benefits and high job satisfaction. When the
income component is greater than 0, the JQI is the average of all four components. As a result, the JQI
can take any values in the [0,1] range. In other words,3
������������ ������������������������������������������������������ = 0 ������ℎ������������ ������������������ = 0
������������������������������������������������������ + ������������������������������������������������������������������ + ������������������������������������������������������������������ + ������������������������������������������������������������������������������������������
������������ ������������������������������������������������������ > 0 ������ℎ������������ ������������������ =
4
2. Findings
The different dimensions of job quality exhibit substantial variation across countries in LAC, especially
in terms of benefits and security (figure 1). Guatemala, Honduras, Bolivia, Paraguay, and Peru have the
lowest levels of job quality in terms of benefits coverage. In other words, workers in these economies
face higher levels of vulnerability during sickness or in old age. In contrast, Uruguay and Chile rank the
highest in terms of this dimension. Honduras, Guatemala, and El Salvador have lower levels of job
security. Cross-country differences in terms of labor income poverty are smaller, with a larger share of
workers in Honduras and Peru earning below the US$6.85-per-day poverty line. Job satisfaction is also
more stable across countries, although it is lowest in those countries that perform poorly in other
dimensions, such as Honduras, Peru, and Guatemala.
3 For more methodological details, please see Brummundi, Mann, and Rodriguez-Castelan (2018).
3
Figure 1 Dimensions of the JQI by Country (Latest Year)
Source: Own elaboration based on SEDLAC (CEDLAS and the World Bank).
Note: The values for Argentina (urban), Brazil, Colombia, the Dominican Republic, El Salvador, and Uruguay for 2023 are based on preliminary data. a/ Values for
Bolivia, Chile, Mexico, Paraguay, and Peru in 2023 are unavailable, hence 2022 data were used. The values for Bolivia for 2022 are based on preliminary data.
b/ Values for Guatemala are from 2014 and for Honduras are from 2017.
When aggregating the different dimensions of the JQI, Chile and Costa Rica have the highest levels of
job quality in the region, while Honduras, Guatemala, and Peru rank the lowest (figure 2). While the
time coverage varies across countries, most of them experienced an increase in job quality between the
first and latest available years of the period of study. Argentina, Colombia, and Peru saw the largest
increases; Guatemala and Honduras, however, saw declines.
Figure 2 Job Quality Index (JQI) (First vs. Latest Year Available)
Source: Own elaboration based on SEDLAC (CEDLAS and the World Bank).
Note: The first year and latest available year are indicated for each country: take, for example, Chile, 2006 (first year) – 2022 (latest year). The values for
Argentina (urban), Brazil, Colombia, the Dominican Republic, El Salvador, and Uruguay for 2023 are based on preliminary data. Table A3 displays the JQI
by country and by year covered. For more details, see The Quality of Jobs.
4
The quality of jobs is highly correlated with welfare indicators across countries. On average, a 0.01
increase in the JQI is associated with about a 0.9 percent increase in GDP per capita (figure 3, panel c).
Improvements in job quality are also linked to reductions in poverty and, to a lesser extent, in inequality.
After controlling for time trends and country time-invariant heterogeneity, the results show that a 0.01
improvement in the JQI is associated with about a 0.9 percentage point decrease in poverty, on average
(figure 3, panel a). The association between inequality and the JQI is negative but weaker (figure 3,
panel b).
Figure 3 Poverty, Inequality, GDP per capita, and Quality of Jobs
(a) Poverty US$6.85(2017 PPP) vs. JQI (b) Gini vs. JQI
(c) GDP per capita (log) vs. JQI
Source: Own elaboration based on SEDLAC (CEDLAS and the World Bank).
Note: The dots are the residuals from a cross-country-year regression of the corresponding variable (either poverty rates, Gini coefficient, GDP
per capita, or JQI) on a set of year and country fixed effects. For panel a the fitted line equation is ������ = −90.778������ − 1.09������ 09 ; for panel b the fitted
line equation is ������ = −15.886������ − 6.22������ 11, and for panel c the fitted line equation is ������ = 0.927������ + 1.17������ −10.
5
In most LAC countries, the elasticity of employment to economic growth strengthens when adjusting
for the quality of the newly created jobs. The link between GDP growth and total employment is often
cited to highlight the importance of economic growth for job creation and poverty reduction. The
employment-to-output elasticity is a summary measure of these linkages. An elasticity estimates
between 0 and 1 indicates that GDP growth is associated with both employment and productivity
growth. An elasticity greater than 1 indicates that GDP growth is linked to job creation but to negative
productivity growth, because employment grows faster than output (Kapsos 2006). Finally, a negative
elasticity implies that economic growth is associated with fewer jobs but higher productivity. Over the
period considered in this technical note, most countries experienced an employment-to-output
elasticity between 0.17 and 0.92, implying that economic growth was accompanied by more jobs and
higher productivity (figure 4). The only country where GDP growth was accompanied by a
disproportionate increase in jobs and therefore lower productivity (an elasticity greater than 1) was
Honduras. Adjusting changes in employment by the changes in the quality of jobs shows that, in general,
not only the quantity but also the quality of jobs increased with economic growth, on average, in 9 out
of the 15 countries considered.4 The most notable exception is Honduras, where the quality-adjusted
elasticity declined to 0.75. This means that economic growth was linked to more jobs, but not
necessarily better jobs, which is consistent with the concurrent decline in labor productivity discussed
above.
Figure 4 Employment-to-GDP Elasticities
Source: Own elaboration based on SEDLAC (CEDLAS and the World Bank).
Note: The values for Argentina (urban), Brazil, Colombia, the Dominican Republic, El Salvador, and Uruguay for 2023 are based on preliminary data. The
elasticities were calculated using the time series of employment and GDP and an OLS regression over the range of years for which the JQI is available. The
dependent variable is the logarithm of total employment and the independent variable is the logarithm of GDP in constant US dollars. The JQI-corrected
elasticities are estimated using the same methodology, but using the logarithm of total employment multiplied by the JQI. Intuitively, if there are no changes in
the JQI over time, then the JQI-corrected elasticity should coincide with the default. In contrast, if the quality of jobs improves (worsens), the JQI-corrected
elasticity should be larger (smaller) than the default.
4 It should be mentioned that, in contrast to the standard employment-to-growth elasticity, the point estimate of the JQI-adjusted
elasticity does not have a direct interpretation. The latter can only be used in conjunction with the former to test whether changes
in quality accentuate or diminish the link between growth and job creation.
6
Differences in labor productivity across sectors and across countries in LAC were strongly correlated
with job quality (figure 5). Countries with higher levels of labor productivity tended to be among those
with higher job quality. Accordingly, the quality of jobs tended to be higher in the industry sector (which
has the highest productivity), followed by services and agriculture (which has the lowest) across most
countries.
Figure 5 Productivity Subindices by Sector vs. JQI (Circa 2023)
Source: Own elaboration based on SEDLAC (CEDLAS and the World Bank).
Note: Labor productivity is measured as value added per worker, shown in logarithm. The fitted line equation is Log(y)= 1.7232x + 2.9429. For more
details on the latest year available from the JQI see table1 (The latest year of the productivity index is 2022, so the same year was used for the JQI; in the
case of Guatemala and Honduras, the year corresponding to the latest year available for the JQI was used.)
The poorer quality of jobs in less-developed economies may reflect their stage in the structural
transformation process. In several countries in the region, rural labor markets are often associated with
casual or seasonal jobs or subsistence agriculture (Morris et al. 2020). Thereby, it is to be expected that
the quality of jobs would be lower than in urban areas. In fact, this is the case for all countries in LAC
except Uruguay (figure 6, panel a).5 The rural-urban gap is the widest in Honduras.
Gender gaps in job quality are pervasive across LAC (figure 6, panel b). When women’s JQI is compared
to men’s JQI, with the former represented as a proportion of the latter, the difference between the two
JQIs (as a percentage) is the gender gap. The gender gaps are the widest in Peru (19 percent), Bolivia (12
percent), and Ecuador (10 percent). The countries with the smallest JQI gender gaps are Panama (5
percent) and Chile (3 percent). Honduras and Guatemala display very small gender gaps as well, but
their levels of job quality for both genders are extremely low when compared to those of other
countries in LAC.
Within-country differences in job quality are also associated with workers’ stages in the life cycle.
Across all countries in LAC, the JQI is the lowest for the youngest age group, that is, those 15 to 24 years
old (figure 6, panel c). The JQI age gap between young and prime-age workers (those 25 to 54 years old),
5 The data for Argentina do not include rural areas.
7
following the method of calculating the gender gap, is the widest in Uruguay (42 percent), Bolivia (37
percent), and Peru (30 percent). At the other extreme are El Salvador and the Dominican Republic with
gaps of about 8 and 14 percent, respectively. The age gap in job quality between prime-age and senior
workers (again, calculated as per the gender gap) is narrower. In most countries, this gap is about 5
percent or lower. Only El Salvador, Mexico, Guatemala, and Honduras have more-significant age gaps in
job quality between senior and prime-age workers that range between 14 and 22 percent. In contrast,
there is a “reversed” age gap in Uruguay and Panama, with the senior group having jobs of higher
quality than prime-age workers (by about 2 to 7 percent).
Figure 6 JQI by Area, Gender, and Age (Latest Year Available)
Source: World Bank staff calculations based on SEDLAC (CEDLAS and the World Bank).
Note: The values for Argentina, Brazil, Colombia, the Dominican Republic, El Salvador, and Uruguay for 2023 are based on preliminary data. Argentina only has
urban coverage. Young: 15–24 years old; prime age: 25–54 years old; senior: 55–64 years old. For more details, see The Quality of Jobs.
8
3. Final Remarks
This technical note describes and provides an overview of a Job Quality Index (JQI) analysis for LAC. It
finds that there is substantial heterogeneity in job quality between and within countries in the region.
Better jobs are strongly associated with the level of economic development and the stage of structural
transformation of a country. Within countries, those with the worst-quality jobs are rural, female, and
young workers.
9
References
Alkire, Sabina, and James Foster. 2011. “Counting and Multidimensional Poverty
Measurement.” Journal of Public Economics 95 (7–8): 476–87.
Brummund, Peter, Christopher Mann, and Carlos Rodriguez‐Castelan. 2018. “Job Quality and
Poverty in Latin America.” Review of Development Economics 22 (4): 1682–1708.
Bruns, A., and N. Pilkauskas. 2019. “Multiple Job Holding and Mental Health among Low-income
Mothers.” Women’s Health Issues 29 (3): 205–12.
Dickey, H., V. Watson, and A. Zangelidis. 2011. “Is It All About Money? An Examination of the
Motives behind Moonlighting.” Applied Economics 43 (26): 3767–74.
Kapsos, Steven. 2006. “The Employment Intensity of Growth: Trends and Macroeconomic
Determinants.” In Labor Markets in Asia: Issues and Perspectives, edited by Jesus Felipe
and Rana Hasan, 143–201. London: Palgrave Macmillan UK.
Klinger, S., and E. Weber. 2020. “Secondary Job Holding in Germany.” Applied Economics 52
(30): 3238–56.
LaMontagne, A. D., L. S. Too, L. Punnett, and A. J. Milner. 2021. “Changes in Job Security and
Mental Health: An Analysis of 14 Annual Waves of an Australian Working-population
Panel Survey.” American Journal of Epidemiology 190 (2): 207–15.
Morris, Michael, Sebastian Ashwini Rekha, Viviana Maria Eugenia Perego, John. D.
Nash, Eugenio Diaz-Bonilla, Valeria Pineiro, David Laborde, Pradeep Prabhala, Joaquin
Arias, Carmine Paolo De Salvo, and Miriam Elizabeth Centurion. 2020. Future Foodscapes
: Re-imagining Agriculture in Latin America and the Caribbean (English). Washington, D.C:
World Bank Group.
http://documents.worldbank.org/curated/en/942381591906970569/Future-Foodscapes-
Re-imagining-Agriculture-in-Latin-America-and-the-Caribbean.
Pouliakas, Konstantinos, and Wieteke S. Conen. 2023. “Multiple Job-holding: Career Pathway or
Dire Straits?” IZA World of Labor: 356. doi: 10.15185/izawol.356.v2.
Watson, B., and L. Osberg. 2018. “Job Insecurity and Mental Health in Canada.” Applied
Economics 50 (38): 4137–52.
World Bank. 2013. World Development Report 2013: Jobs. Washington, DC: The World Bank.
Zangelidis, A. 2014. “Labour Market Insecurity and Second Job-holding in Europe.” Available at
SSRN 2615268
10
Appendix
Table 1 Detailed Description of the Specific Survey, Years, and Collection Used
Note: Only data for years that are methodologically comparable over time were used. For more details, see comparability dashboard.
The values for Argentina, Brazil, Colombia, Dominican Republic, El Salvador, and Uruguay for 2023 are based on preliminary data.
Table 2 Variables Included in the JQI
11
Table 3 Country-year-level Variables in the JQI
12
13