Stratification and Inequality in Labor Incomes in Spain, 1980–2014; an Analysis by Sectors, Skills, and Gender

Abstract

Intra-country inequality is generally higher now than it was in the 1980s, especially within developed countries. This higher inequality, whether it manifests in income or in the sphere of consumption, is strongly related to social stratification. To analyze these increases in inequality and the resulting social stratification in Spain for the period 1980-2014 (further divided into three sub-periods), we use an input-output framework. This will allow us to apply a Structural Decomposition Analysis (SDA) to annual changes in value added, decomposed into several effects (wages, labor, and consumption), which can be linked to traditional measures of inequality (Gini indexes). First, we find that the labor share of income fell from 1980. Second, the sectoral structure of the labor market is a determinant factor of labor incomes stratification. Third, inequality by gender now is less accentuated than in 1980. Finally, domestic consumption by income quintiles has lost weight in value-added generation.

KEYWORDS: :

• Inequality
• input-output
• stratification
• structural decomposition analysis

JEL CODES: :

• D31
• O41
• P36

Data Availability Statement

The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.

Notes

1 Used in several social studies, such as in Son (2022).

2 Input-output models have been previously used in inequality studies, such as in Parteka and Wolszczak-Derlacz (2015), but not the SDA, up to our knowledge.

3 As we aggregate tables into 8 sectors, and we distinguish 3 levels of skills and 2 categories of gender, we end up having 48 groups associated to labor incomes.

4 Eurostat (2008) offers four alternatives or models, usually named as A, B, C, and D.

5 PS: Primary sector, ES: Energy sector, HT&MHT: High and Medium-high Technology Industry, MLT: Medium-low Technology Industry, LT: Low technology Industry, C: Construction, HTS: High Technology Services, RS: Remaining services.

6 We are fully aware of the different methodologies used in the two surveys. However, we believe that the observed trends in consumption inequalities are not affected by this methodological change, and so it does not affect our results in any significant way.

7 As noted by Dietzenbacher and Los (2000), there is a dependency problem between the empirical data for Δc and ΔA, which is not relevant for our analysis nonetheless. For a possible solution, see Sánchez-Chóliz et al. (2021).

8 It should be remarked that the definitions of productivity are multiple. In particular, those concerning fixed capital productivity can differ, in the short run, from those of labor productivity. Here we consider that increases in productivity are associated with labor and capital savings.

9 See Annex for clarifications on the formulas used for calculating Gini indexes.