Gender discrimination in female-dominated jobs

According to a report from the Italian Vocational Training Development Agency, gender pay gap estimates are lower when differences in job search patterns and job preferences are taken into account. Women search for jobs offering more opportunities for flexible working time even if they are less paid, whereas men pay more attention to the remuneration. Nonetheless, high levels of discrimination may still be found, in particular among women educated to primary or third level.

Different job search patterns

The study Gender wage gap and typically female-dominated jobs (Differenziale salariale di genere e lavori tipicamente femminili (1.2Mb PDF)) is part of a conclusive report of a three-year project carried out by the Italian Vocational Training Development Agency (Istituto per lo Sviluppo della Formazione Professionale dei Lavoratori, Isfol) on behalf of the Ministry of Labour, Health and Social Policy (Ministero del Lavoro, della Salute e delle Politiche Sociali). The report is based on the 2006 wave of the Participation, Labour, Unemployment Survey, Isfol PLUS (in Italian, 253Kb PDF) (see IT0611049I for details on survey methodology). It aims to quantify the discriminatory component of the gender wage gap by taking into account two biases that often occur in a gender pay gap estimation.

  • The sample selection bias indicates that the gender pay gap can be affected by women’s and men’s different participation levels in the labour market. The use of the employed population as a survey sample can contain a selection bias: while most men are employed, showing minimal differences from the total population, the employment rate of women is far lower (46.6% in 2007). Thus, significant differences from the total female population could be expected.
  • The endogeneity bias indicates that the wage gap could depend on the type of job and occupation that women and men generally have. This is known to be strongly related to the job characteristics favoured by each of them: while men usually prioritise earnings, women consider other issues, such as the adaptability of working time arrangements.

Main results of study

Both men’s and women’s salaries increase with education, age, work experience and job tenure; however, the increase is almost always greater for men than for women. The only exception is with regard to job tenure. When introducing corrections for both the sample selection and endogeneity biases, there is no impact on men’s earnings, which means that their compensation would not change regardless of whether they worked in male or female-dominated occupations. However, women employed in female-dominated jobs earn less than those working in non-female-dominated jobs.

The hourly gender pay gap is 8.75%. It can be subdivided into an ‘endowment effect’ and a ‘coefficient effect’ (see table). The former captures the difference between women and men in terms of qualifications, competences, work experience and education, while the latter measures the difference in compensation between men and women having the same characteristics.

Hourly gender pay gap, endowment and coefficients effects according to selection processes (%)
  Hourly gender pay gap Endowment effect Coefficient effect (discrimination)
Without bias correction 8.75 -6.66 15.41
With bias correction 8.75 -2.08 10.83

Source: Centra and Cutillo, 2009

When neither bias is taken into account, the pay discrimination decreases from 15.4% to 6.66% as a result of the endowment effect. These figures are similar to the findings of Pissarides et al (2005), who reported respective proportions of 16% and 8.5%. However, they differ slightly from the results of Addabbo and Favaro (2007), who reported proportions of 18% and 6%. Both of those studies were based on European Community Household Panel (ECHP) data but used different estimate methods without any bias correction. When the two biases are taken into account, the pay discrimination declines from 10.83% to 2.08%.

The discriminatory effect decreases as the person’s educational level increases, from 20.96% for those having primary education only to 5.45% for those having upper secondary education – corresponding to ISCED 3 level, according to the International Standard Classification of Education (ISCED). However, the discriminatory effect increases to 12.09% for those with a university qualification. This shows the persistence of a ‘glass ceiling’ effect, thus confirming results from the 2005 PLUS wave (IT0611049I).


Following a comparison of the different sources available in Italy (Rustichelli, 2007), the latest study offers a more accurate methodology to detect the discrimination component of the gender pay gap by taking into account the actual labour market situation of women. This issue is particularly relevant in the Italian labour market – especially in the country’s southern regions – since women’s employment rate is far below the target set in the Lisbon Strategy of the European Union.

Nonetheless, differences in job search attitudes between men and women explain only part of the wage discrimination that women suffer. Female-dominated occupations include a large proportion of jobs where women are paid less and report higher unemployment rates. Thus, it appears that employers are still apprehensive about maternity leave, as well as women’s greater interest in part-time jobs and, in general, lower availability to work long hours when required. Such reservations reflect ‘the persistence, and sometimes the tightening, of gender-related competences inside organisations’, as highlighted by Catemario et al (2008). They also reveal employers’ tendency to monitor their employees according to the ‘face time’ that they spend at work rather than on their actual performance, thus discriminating against women both with regard to rewarding their effort and in recruitment decisions.


Addabbo, T. and Favaro, D., ‘Differenziali salariali per sesso in Italia. Problemi di stima ed evidenze empiriche’ [Gender wage gaps in Italy. Problems in estimation and empirical results], in Rustichelli, E. (ed.), Esiste un differenziale retributivo di genere in Italia? [Does a gender wage gap exist in Italy?], European Social Fund Books, Rome, Isfol, 2007.

Catemario, M.G., Merelli, M. and Ruggerini, M.G., Differenziali retributivi di genere e organizzazione del lavoro. Una indagine qualitativa (Zip file) [Wage differentials and work organisation. A qualitative investigation], Rome, Isfol, 2008.

Centra, M. and Cutillo, A., ‘Differenziale salariale di genere e lavori tipicamente femminili’ (1.2 Mb PDF) [Gender wage gap and typically female-dominated jobs], Isfol Studies 2009/2, 2009.

Pissarides, C., Garibaldi, P., Olivetti, C., Petrongolo, B. and Wasmer, E., ‘Wage gaps’, in Boeri, T., Del Boca, D. and Pissarides, C. (eds), Women at work. An economic perspective, Oxford University Press, 2005.

Rustichelli, E. (ed.), Esiste un differenziale retributivo di genere in Italia? [Does a gender wage gap exist in Italy?], European Social Fund Books, Rome, Isfol, 2007.

Mario Giaccone, Cesos

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

Eurofound welcomes feedback and updates on this regulation

Aggiungi un commento