Volume 44, Number 174,
July-September 2013
Salary Gaps in Uruguay: Gender, Segregation
and Unequal Labor Qualifications
Alma Espino
Job Skill Mismatch ( ...continuation )

Where hi is the percentage of men working in occupation i and mi is the percentage of women working in occupation i. This index varies between 0 and 1, taking 0 as the value when the occupational distribution between men and women is identical, and 1 when men and women do not overlap in any occupation, which would be perfect occupational segregation. This index is normally interpreted as the proportion of occupied women (or men) that would be necessary to change jobs in order to achieve perfect integration or uniform distribution. The di can be calculated for types of jobs or sectors of activity, sensitive to the level of aggregation of the categories being analyzed. In other words, at higher aggregation, the segregation reached by the index is lower. However, besides the level, the evolution over time may also differ when comparing between areas.3 In order to analyze the evolution of the di, the researcher must take into account whether the changes verified are statistically significant. To do so, confidence intervals are constructed using the bootstrapping technique,4 which allows for the estimation of the distribution of statistics based on resampling with reposition, in order to construct confidence intervals for the statistic.

Measuring Skill Mismatch

Commonly used statistical methods to measure skill mismatch are based on the effectively observed correspondence between jobs, occupations and levels of education with respect to the classification of occupations, at the observed average distribution of the occupied persons (Verdugo and Verdugo, 1989) or at the mode of the distribution. The information used as a reference for worker qualifications is the level or years of education. To understand the features of job positions according to required level of education – in the absence of specific surveys – an occupation classification is normally used. These, however, tend not to be updated (Hartog, 2000). In addition, converting the classification scales to years of education is not always clear (Halaby, 1994), there is an implicit assumption that all employees with the same level of education are perfect substitutes and there are set educational level requirements for a type of occupation (Chevalier, 2003). All of these aspects weaken the analysis of mismatch.

This study considers over-qualified workers as those with more years of formal education than are required for the job they hold, according to isco-885 , and under-qualified as those with fewer years of education than required for their classification. The imposition of skill and qualification levels for occupations (isco-88) makes it possible to contrast the qualifications required by certain jobs and the skills held by the persons in these occupations. It is therefore a standard criteria based on pre-established formal requirements. Espino (2011) indicates that this method allows for more appropriate analysis of the Uruguayan case than the method of average or mode, because the distribution observed in labor qualifications does not support the use of the average or mode as valid references. Reference values do present some objectivity in defining requirements and are independent of the qualifications effectively observed for the workforce of the country.6

Salary Differences and Segregation

The methodology proposed by Bayard et al. (2003) was followed to include gender segregation among the explanatory variables of salary differences between men and women. To do so, the assumption was that these differences are a function of the individual characteristics of human capital and those of the feminization of the labor environment for each worker, represented by the percentage of women in each occupation, sector and size of establishment.

3 Greater disaggregation occurs when the sample size available is not very large. Due to the lack of representation, there may be some occupations with zero observations out of the total, especially for women in positions that are mainly filled by males, which distorts the level of segregation found.

4 Bootstrapping calculations did not use any prior assumption regarding the distribution, they corrected for bias in the average estimated value and were calculated for 95% confidence intervals. One hundred repetitions of the original sample were carried out.

5 The International Standard Classification of Occupations (isco88) has been used since 2000 by the National Statistics Institute (ine) to categorize the types of tasks that employed persons carry out in Uruguay. It uses four levels or skills groups defined by the categories and levels of the International Standard Classification of Education (isced). Skill levels are not specified within large occupational areas, but they are directly associated with the major groups of occupations. Eight of the large groups of occupations are directly linked to a skill level. For example, the group of scientific and professional occupations requires the highest skill level, and the group of non-qualified occupations requires the lowest skill level.

6 Similar to other studies, the results for Uruguay are maintained in comparative terms even if the method of the average is used, but the magnitudes are very different, because nearly three-quarters of the occupied persons end up as highly qualified. When the analysis is done using the mode, the same occurs for the relative values, but the magnitudes are somewhat closer to the standard method (Espino, 2011). For further discussion of this, see G. Quintini (2011).

Published in Mexico, 2012-2017 © D.R. Universidad Nacional Autónoma de México (UNAM).
PROBLEMAS DEL DESARROLLO. REVISTA LATINOAMERICANA DE ECONOMÍA, Volume 49, Number 193, April-June 2018 is a quarterly publication by the Universidad Nacional Autónoma de México, Ciudad Universitaria, Coyoacán, CP 04510, México, D.F. by Instituto de Investigaciones Económicas, Circuito Mario de la Cueva, Ciudad Universitaria, Coyoacán,
CP 04510, México, D.F. Tel (52 55) 56 23 01 05 and (52 55) 56 24 23 39, fax (52 55) 56 23 00 97, www.probdes.iiec.unam.mx, revprode@unam.mx. Journal Editor: Moritz Cruz. Reservation of rights to exclusive use of the title: 04-2012-070613560300-203, ISSN: pending. Person responsible for the latest update of this issue: Minerva García, Circuito Maestro Mario de la Cueva s/n, Ciudad Universitaria, Coyoacán, CP 04510, México D.F., latest update: June 27th, 2018.
The opinions expressed by authors do not necessarily reflect those of the editor of the publication.
Permission to reproduce all or part of the published texts is granted provided the source is cited in full including the web address.
Credits | Contact

The online journal Problemas del Desarrollo. Revista Latinoamericana de Economía corresponds to the printed edition of the same title with ISSN 0301-7036