Increase in generational wage gap since early 1990s

According to a recent study carried out by the Bank of Italy, entry wage levels of young male employees show a significant decline since the early 1990s. Despite a noticeable increase in educational levels, no changes emerge in the ‘age-earnings profile’ over the past three decades. These findings reflect both a productivity slowdown of the Italian labour force and a decline in aggregate wage share among workers.

In September 2007, the Research Unit of the Central Bank of Italy (Banca d’Italia) published a working paper entitled The generation gap: Relative earnings of young and old workers in Italy (862Kb PDF). The paper examines the developments of labour market entry wages, as well as the gap between entry wages and wages of older workers, over the period 1975–2004.

About the study

The Bank of Italy study draws on data from two different sources: a sample of 70,000 workers taken from the administrative records of Italy’s National Social Security Institute (Istituto nazionale di previdenza sociale, INPS) and the Survey on household income and wealth (SHIW) (Indagine sui bilanci delle famiglie italiane) which is carried out biennially by the Bank of Italy since the mid 1960s.

INPS data

INPS records data on the occupational status, qualification, gender and age of employees working in the private sector, with the exception of agriculture. INPS data does not provide any information on previous periods of employment under a different status – such as self-employed, freelance or undeclared worker – or on the educational level of employees. In order to overcome the lack of information on the educational level of employees, a proxy of this variable is used by assuming that workers who entered the labour market at the age of 21–22 years have most probably completed secondary education, while those entering the labour market aged 25–26 years are likely to have attained third-level education and hold a university degree.

SHIW data

The study on the generation gap also draws on data from the SHIW in order to control for socio-demographic variables which are not available from the INPS administrative records. The SHIW gathers data on the incomes and savings of Italian households. The most recent surveys comprised each a population sample of about 8,000 households, covering 24,000 individuals distributed over about 300 Italian municipalities. Each survey includes a full section on employment and income (see questionnaires). Over the years, the scope of the SHIW has expanded, comprising today information on wealth and other aspects of households’ economic and financial behaviour (see questionnaires), as well as a section on employment and income.

Trends of labour market entry wages

The Bank of Italy analysis on the relative earnings of young and older workers only looks at the earnings of male workers over the last three decades.

In terms of entry wages, both groups of young workers (aged 21–22 years and 25–26 years) report important improvements up to the early 1990s; however, a significant deterioration in levels of entry wages was seen from 1992 onwards (see Figures 1 and 2). In the late 1990s, the development of labour market entry wages begins to differ between the two groups, depending on their educational level.

Employees, who entered the labour market at the age of 21–22 years and thus have a lower educational level, report an increase in their initial wages of 35% over the period 1976–1992 and then a decline of 12% in subsequent years until 2004 (see bold line in Figure 1). This decline in entry wages for the lower educated group of employees is not compensated by steeper age-related earnings profiles as illustrated by the thinner lines in Figure 1.

Entry wages and career development of young workers when entering the labour market aged 21–22 years, 1976–2004

Entry wages and career development of young workers when entering the labour market aged 21–22 years, 1976–2004

Notes: Wages are deflated with the consumer price index. Index: 1976=1.

The thick line highlights labour market entry wages of subsequent cohorts, while the thinner lines represent the wage profiles over the working life, namely the age-earnings profile.

Source: Bank of Italy, September 2007, p. 18

For the higher educated group of employees – aged 25–26 years when entering the labour market – entry wages increased by 37% over the period 1976–1992 (see bold line in Figure 1). Compared with the lower educated group of employees, this represents a two percentage point difference. In the following decade, the labour market entry wages of employees with a higher educational level declined by 8%, while their level recovered slightly in 2004. This brings the overall decline in entry wages for this group to 4% relative to the 1992 peak.

Entry wages and career development of young workers when entering the labour market aged 25–26 years, 1976–2004

Entry wages and career development of young workers when entering the labour market aged 25–26 years, 1976–2004

Notes: Wages are deflated with the consumer price index. Index: 1976=1.

The thick line highlights labour market entry wages of subsequent cohorts, while the thinner lines represent the wage profiles over the working life, namely the age-earnings profile.

Source: Bank of Italy, September 2007, p.18

Generational wage gap

The wage gap between younger workers aged 19–30 years and older workers aged 31–60 years declined from about 25% to 13% in the period between the mid 1970s and the late 1980s. It then increased sharply showing wage differentials of up to over 30% in the early 1990s. This was followed by a more moderate increase in the wage gap, while attaining the highest level in the early 2000s (see continuous line in Figure 3). In 2004, older workers earned 35% more than younger workers, compared with a wage gap of less than 20% in 1989. When weighing the data to control for compositional differences, the overall trend in the gap between wages of younger and older employees is replicated (see marked line in Figure 3).

Net monthly wage gap between younger and older employees, 1977–2004 (%)

Net monthly wage gap between younger and older employees, 1977–2004 (%)

Notes: Continuous line = net monthly wage differential; marked line = net monthly wage differential, weighted data.

Younger employees aged 19–30 years; older employees aged 31–60 years. Net wage differential logarithm based on SHIW data: self-reported information on net annual income and months worked; from 1987, SHIW also collected information on working hours.

Source: Bank of Italy, September 2007, p. 20

Impact of educational and demographic changes on wages

Educational and demographic changes are two important factors affecting the analysis of the generational wage gap: the number of younger male employees aged 19–30 years as a share of the population aged 19–60 years has declined from 30% in 1977 to about 20% in 2004, while the educational level of employees has increased over time. The proportion of secondary school graduates increased from less than 13% to 44% of the labour force in the period between 1975 and 2004; the proportion of university graduates also increased from 4% to 14% over the period under examination.

To verify whether these changes may account for the wage gap between younger and older employees in the 1990s, the authors estimated the average wage of employees of a certain age group in relation to that of employees of another age group. The table below outlines these estimates of relative wages, taking 1993 as the reference year for entry to the labour market. Column 1 outlines the estimates pooling relative wages for all educational categories: in this case, compared with the reference group (51–60 years), the wage gap widened from around 42% in 1993 to 55% in 2004 for those aged 21–25 years, and from 27% (1993) to almost 40% (2004) for those aged 26–30 years; the gap thus increased by about 13 percentage points for both age groups. The increase in wage differentials does not change significantly when taking into account educational attainment (column 2) and different substitution rates among workers with different educational levels (column 3). In particular, the latter estimate shows that workers holding a university degree can be replaced with those having a lower educational level, without any significant loss in productivity.

A breakdown of relative wage estimates by educational level is given in columns 4 to 6, indicating that young workers holding a university degree are those showing the highest wage gap with older employees (51–60 years): 26.7% for employees in the 21–25 age group and 20.1% for those aged 26–30 years.

Estimated relative wages, by age and educational level
This table outlines the estimates of relative wages across educational categories, taking the 51–60 age group as a reference group and 1993 as a reference year for labour market entry.
  Relative wages for all educational categories Educational attainment: compulsory and secondary education Substitution rates across educational levels Compulsory education Secondary school Third-level education
  (1) (2) (3) (4) (5) (6)
Age-education specific relative supply (ERS) -0,004 0.011   0.035 -0.140** 0.021
ERS* compulsory education     -0.092      
ERS* secondary school     -0148**      
ERS* third-level education     0.022      
21–25 years -0.551** -0.614** -0.629** -0.310** -0.560** -0.685**
26–30 years -0.397** -0.476** -0.475** -0.191** -0.399** -0.552**
31–35 years -0.228** -0.309** -0.294** -0.066** -0.249** -0.324**
36–40 years -0.112** -0.193** -0.175** -0.011 -0.145** -0.140**
41–45 years -0.094** -0.176** -0.157** 0.029 -0.078* -0.190**
46–50 years -0.024 -0.104** -0.102** 0.020 -0.036 -0.044
21–25 years *pre-1993 0.134** 0.130** 0.152** 0.074** 0.083 0.267**
26–30 years *pre-1993 0.126** 0.124** 0.133** 0.098** 0.112* 0.201*
31–35 years *pre-1993 0.017 0.015 0.009 0.033 0.048 -0.007
36–40 years *pre-1993 -0.002 -0.005 -0.014 0.023 0.048 -0.007
41–45 years *pre-1993 0.042 0.041 0.031 0.027 0.018 0.080
46–50 years *pre-1993 0.012 0.011 -0.000 0.032 0.005 -0.017
Compulsory education   0.211** 0.162**      
Secondary school   0.037 0.059**      

Notes: Reference age group is 51–60 years; relative supply only includes men of age-educated cell; sample years: 1984–2004. significant at 10%; * significant at 5%; ** significant at 1%.

Source: Bank of Italy, September 2007, p. 26

Commentary

According to the Bank of Italy analysis on the generational wage gap, the development in labour supply and technological change, as well as the selection of less productive young workers to enter the labour market, are not plausible explanations for the observed pattern in the wage gap between younger and older workers. The trend of a widening generational wage gap may be due to increasing labour market flexibility in terms of hiring young people who enter the labour market. Such flexibility has been achieved through the introduction of new non-standard employment contracts aimed at cutting substantially social security contributions – such as the ‘training and work’ employment contract in 1984 and the new apprenticeship scheme in 1998 – or meeting temporary labour shortages by introducing temporary agency work in 1998.

Other reasons for such a wage gap between older and younger workers may be ascribed to the extensive use of work contracts based on self-employment, such as freelancers and partners in social cooperatives, and a rising proportion of new labour market entrants working less than three and six months a year. In addition, young workers holding a university degree report the worst relative wage performance. This finding reveals that changes in the educational level of young workers did not match the developments of labour demand, since the tasks that higher educated employees perform can be carried out by less educated workers without any significant productivity loss. This partly confirms the findings of the 2002 Quality of work survey (in Italian) by the National Training Agency (Istituto per la Formazione dei Lavoratori, Isfol) on the perception among younger educated workers that skill levels are in excess of those required for their job (see an outline of the findings in the 2005 EWCO Survey Data Report on Quality of work in Italy survey, 2002).

Mario Giaccone, Fondazione Seveso

Useful? Interesting? Tell us what you think. Hide comments

Eurofound welcomes feedback and updates on this regulation

Add new comment