Gender gap in reservation wages and the choice of education field

ABSTRACT

I examine the gap in reservation wages between men and women who enter the labour market for the first time and link it to the choice of education field by gender. Using data from 2006 to 2015 Labour Force Survey for Poland, I show that men and women with the same education value their qualifications differently, which points to behavioural differences by gender. The gender difference in returns to education fields is found to be the single most important predictor of the gender gap in reservation wages. Women educated in mathematics, IT and physics are found to report higher reservation wages than other women and there is no difference in pay expectations between them and men. The results contribute to the discussion on the importance of a higher presence of women in STEM fields as a way of achieving greater gender equality in the labour markets.

KEYWORDS:

• Gender wage gap
• reservation wages
• sex segregation in education
• gender inequality

JEL CLASSIFICATION:

• J16
• J31
• J24

I. Introduction

Despite the years of efforts and the number of equal pay regulations gender wage inequality prevails and women earn less than men in nearly all developed countries. Existing research shows that the fact that women earn on average less than men may be explained by a number of factors. They relate to gender difference in labour market experience and education (Becker 1971; Mincer and Polachek 1974), the childbearing responsibilities that are mostly placed on women (Angelov, Johansson, and Lindahl 2016; Cukrowska-Torzewska and Lovasz 2016, 2020), the choice of job, sector, or working hours (Blau and Kahn 1997, 2000; Goldin 2014).¹ Recent studies show that behavioural factors, such as risk aversion and negotiation skills, are not less important (Manning and Swaffield 2008; Greig 2008; Le et al. 2011; Cobb-Clark and Tan 2011). However, even if these factors are taken into account and we compare wages of ‘similar’ men and women, the gender wage gap is still present.

Given the enormous amount of research on the topic, surprisingly little attention is paid to gender differences in wage expectations, which are key to wage negotiation process and the realized wage. Contributions along this line include Caliendo, Lee, and Mahlstedt (2017), who link the gender gap in realized wages to the gap in reservation wages, showing that the latter explains about half of the former. The gap in reservation wages between men and women is thus an important, yet largely neglected by researchers, source of gender wage inequality.

Some research that relates to the gender gap in pay expectations and reservation wages exists but it mostly focuses on the size of the gaps and their sources (e.g. Säve-Söderbergh (2019) for Sweden, Filippin and Ichino (2005) for Italy, Zambre (2018) for Germany, Brown, Roberts, and Taylor (2011) for the UK). Despite that reservation wage, defined as the lowest wage one would consider accepting for a job, and expected wage are not the same, they remain related as wage expectations affect the level of the reservation wage (Brown and Taylor 2013).

With this paper I contribute to our understanding of sources of gender wage inequality by examining men’s and women’s reservation wages in relation to their education field. Previewing the main findings, I find that reservation wages of men and women differ by education field and the type of education accounts for most of the gender gap in reservation wages. The lowest and insignificant gap in reservation wages is seen among men and women educated in mathematics, IT and physics, which is due to women’s increased reservation wage level. The results contribute to the discussion on the importance of a higher presence of women in STEM fields as a way of achieving greater gender equality in the labour markets.

II. Materials and methods

I use 2006–2015² Labour Force Survey (LFS) for Poland, in which all individuals that are seeking a job are asked the following question: ‘What would be the minimum monthly pay for which you would agree to start a job’.³ The answer to this question is consistent with the definition of the reservation wage. Because reservation wages likely differ by previous work experience, the sample is restricted to people who have not worked before. Long-term unemployed, i.e. unemployed for more than 12 months, are dropped because they may adjust their reservation wages in response to limited employment prospects. The results are robust with respect to other definitions of long-term unemployment, e.g. 6, 18 or 24 months. Finally, the sample consists of people with at least secondary education that is not general (i.e. high school) because only for them the field of education can be identified.

The magnitude and sources of the gender gap in reservation wages are uncovered using Blinder (1973), and Oaxaca (1973)decomposition. The decomposition requires that the researcher assigns the reference or so-called ‘non-discriminatory’ pay structure, which will serve as a comparison of the group-specific pay structures. Because in the case of reservation wage it is unclear what should be the reference pay structure, three variations of the decomposition are used: (1) that uses coefficients from the equation for men as the reference coefficients, (2) that uses coefficients from the pooled model as the reference coefficients, and (3) that relies on a weighting the coefficients by the proportion of observations in the two groups, also known as Cotton (1988) extension.

The decompositions are based on regressing male and female logged reservation wage on age, education level (secondary or university), education fields, marital status, months of job search, place of living measured by region and number of inhabitants grouped into four categories, and time-fixed effects. Eight groups of education fields are defined in line with the available information in the survey; they are listed in Table 2.⁴

The gender difference in the returns to education fields is further assessed by a regression analysis, in which logged male/female reservation wage is regressed on individual characteristics and a given group of education fields (i.e. a dummy variable equal to one for a given education field and zero for all other fields), controlling for the variables listed before.

III. Results

Figure 1 displays the size of the gender gap in reservation wages for the selected sample along with the estimates of the gender wage gap for the working population. The mean gender difference in reservation wages amounts to −8%, meaning that women are willing to accept by 8% lower pay than men. This gap is less than half of the gap between working men and women, which in LFS data amounts to −19% and which is comparable to existing studies for Poland (Magda and Cukrowska-Torzewska 2019; Goraus and Tyrowicz 2014).

Figure 1. The size of the gender gap in reservation wages in relation to the gender pay gap

Note: The vertical lines represent 95% confidence intervals.

Table 1 reports decomposition results. The results indicate that the unexplained portion of the gap that is attributed to the gender difference in expected returns to educational field accounts for 52%–80% of the gap – depending on the choice of the reference model. The difference in returns to education fields is the single most important predictor of the gap, irrespective of the model.

Table 1. Gender gap in reservation wages: decomposition results

Total gap	−0.082
	Explained	% of the total gap	Unexplained	% of the total gap
Men’s coefficients as reference
Age	0.007	−8%	−0.042	51%
Education	0.033	−40%	−0.026	31%
Marriage	0.003	−4%	−0.006	8%
Time of job search	−0.000	0%	−0.002	3%
Field of education	−0.011	13%	−0.066	80%
Rest	0.001	−1%	0.027	−33%
Total	0.033	−40%	−0.115	140%
Pooled coefficients as reference
Age	0.004	−4%	−0.039	48%
Education	0.031	−38%	−0.024	29%
Marriage	−0.007	9%	0.004	−5%
Time of job search	−0.000	0%	−0.002	3%
Field of education	−0.034	41%	−0.043	52%
Rest	0.000	−1%	0.028	−34%
Total	−0.006	7%	−0.076	93%
Reweighted coefficients as reference (Cotton)
Age	0.005	−6%	−0.041	50%
Education	0.031	−38%	−0.024	29%
Marriage	0.001	−1%	−0.004	4%
Time of job search	−0.000	0%	−0.003	3%
Field of education	−0.007	8%	−0.070	85%
Rest	0.001	−1%	0.027	−33%
Total	0.031	−38%	−0.113	138%

Table 2 displays the expected returns to education fields by gender and respective gaps in reservation wages. The results confirm what was found using decomposition, namely that there are notable gender differences in the expected returns to education fields. Among women, the highest reservation wages are declared by those educated in mathematics, IT and physics. Women educated in these fields report by 6% higher reservation wage than other women. For men, there is no significant effect, which leads to virtually no gender gap (compare column 3 Table 2).

Table 2. Gender gaps in reservation wages and male and female ‘expected returns’ by education fields

Field of education	N (total)	Gender gap in reservation wages by education type^a	‘Expected returns’ in relation to other fields^b:
			Women	Men
Social care, health	279	−0.116***	0.023	0.004
		(0.043)	(0.020)	(0.037)
Education and pedagogy	394	−0.063*	−0.056***	−0.101***
		(0.037)	(0.018)	(0.033)
Social science, economics, law	2,041	−0.131***	−0.004	0.010
		(0.016)	(0.009)	(0.015)
Arts and humanities	367	−0.121***	0.027	0.036
		(0.034)	(0.023)	(0.028)
Services	1,901	−0.096***	0.006	−0.015
		(0.015)	(0.010)	(0.013)
Agriculture and veterinary	450	−0.091***	−0.037*	−0.059***
		(0.031)	(0.021)	(0.019)
Mathematics, IT and physics	585	−0.036	0.060**	−0.019
		(0.030)	(0.024)	(0.016)
Engineering, production processes, construction	3,339	−0.138***	−0.003	0.029***
		(0.014)	(0.012)	(0.010)

Notes: Clustered standard errors in parentheses. *** denotes p < 0.01, ** denotes p < 0.05, and * denotes p < 0.1. All regressions control for age, education level, marital status, months of job search, size of the place of residence, regions and year dummies. ^a The gender gap is obtained from estimating regressions in which the logarithm of reservation wage is regressed on female dummy. ^b The expected returns to fields of education are obtained from estimating regressions in which logarithm of reservation wage is regressed on a given field of education (i.e. a dummy variable equal to one for a given field and zero for all other fields).

Men in turn have highest reservation wages once educated in engineering, production processes and construction: they declare by 3% higher reservation wage than other men. These fields have the lowest share of women (25%) meaning that they are largely dominated by men. There are no other fields for which men’s expected returns are significantly higher. Women with such qualification do not have significantly different reservation wages than other women. This large discrepancy by gender leads to as high as 13.8% gender gap.

The analysis also shows that the lowest reservation wages are among men and women educated in education and pedagogy, and agriculture and veterinary. The negative impact is stronger for men than for women with the consequence of lower gender inequality.

IV. Discussion

I examine reservation wages of men and women by education fields using LFS data for Poland. The lowest and insignificant gap in reservation wages is seen among men and women educated in mathematics, IT and physics, which is due to women’s increased pay expectations relative to other women.

The results add to the current debate on the need of increasing women’s participation in STEM fields that in the era of digitalization are considered as key for future labour market prospects and economic development (European Commission 2018). It has been argued that an increased presence of women in such fields should have positive impact on their labour market position measured by employment opportunities and wages that would contribute to the reduction of the gender wage gap (Blau and Kahn 2017). The obtained results for Poland do confirm that women educated in mathematics, IT and physics have higher reservation wages than other women. More research that utilizes longitudinal perspective is, however, needed in order to assess whether women’s reservation wages and expectations translate into higher wages – at the market entry and later, once they gain experience. Further research is also needed to better understand the sources of gender gaps in reservation wages – while this study documents gender gaps in reservation wages by education fields, it does not provide evidence on the mechanisms that drive these gaps. There are several possible explanations of the finding that women educated in mathematics, IT and physics have higher reservation wages than other women and that the reservation wage gap between them and men is negligible: e.g. higher relative productivity, increased wage expectations driven by higher wages paid to employees with such qualifications, and self-selection based on individual abilities.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the National Science Centre in Poland grant [2018/31/B/HS4/01562].

Notes

¹ For a comprehensive review of research on the gender wage gap and its sources see e.g. Blau and Kahn (2017), Matysiak and Cukrowska-Torzewska (2021).

² Data for more recent years are not used due to the change in education fields coding.

³ Individuals are asked about monthly pay rather than hourly wage / annual earnings because monthly pay is the most common way of reporting earnings in Poland.

⁴ The „life sciences” are dropped from the analysis due to the insufficient number of observations that belong to this education field.