Chiara Moruzzi, Francesca Martinoli & Federico Taddei
Final assignment for the “Data Analysis” course in the Public and Corporate Communication Master’s program, curriculum in Data Analytics for Politics, Society and Complex Organizations, 2022-2023, University of Milano. Instructor: Giulia Dotti Sani
Introduction, literature review, and hypotheses
In this short analysis, we use Eurobarometer survey data (Round 87.4 – June 2017) to investigate how gender affects the propensity to salary negotiation.
Gender is one of the most investigated factors regarding the propensity to salary negotiation: many studies have been conducted, sometimes with conflicting results. Overall, studies indicated differences between men and women in starting salary negotiation (Johnson, 2016; Hannah Riley Bowles et al., 2007; Babcock & Laschever, 2003) and in the various strategies used to influence their employers (Lauterbach & Weiner, 1996).
However, less is known about whether country-level characteristics account for the gender gap in salary negotiation. Thus, we further test whether the gender gap in salary negotiation varies across European countries with different levels of gender equality measured through the Gender Equality Index (GEI) developed by the European Institute for Gender Equality. In particular, we use the values of the index designed to capture gender equality in the work domain.
We formulate and test the following hypotheses:
• H1: Women are less inclined to negotiate their salary.
• H2: The gender gap in salary negotiation is smaller in countries with higher levels of societal gender equality.
Our results confirm a difference in salary negotiation between genders. In contrast, we found no support for H2. In fact, our findings indicate that the gender gap in salary negotiation is larger in countries with higher scores of societal gender equality
Data and Method
The dataset we chose (Eurobarometer 87.4 – 2017) is focused on Gender Equality. The target population includes residents of all 28 European countries (the survey was done before BREXIT); the randomization method used is multistage sampling.
The first thing we did was to recode the dependent variable (Comfortable With Negotiating Salary) into a dichotomous variable: we grouped the original categories “Not very comfortable” and “Not at all comfortable” into “No” and the categories “Very comfortable” and “Fairly Comfortable” into “Yes”. We assigned missing values to all the respondents to which the question does not apply (mostly independent workers who, by definition, do not have to negotiate their salary).
Our main independent variable at the individual level is gender, which we left dichotomous as it was.
For GEI scores we created a specific variable with the corresponding value for each country using 2017 data. This index considers work domains such as participation, sectorial segregation, work-life balance, and career prospects. It is measured on a scale between 0 and 100 points, where 100 is the maximum and indicates full gender equality.
The variable’s distribution is shown in Figure 1. The European mean value is 70.6, displayed by the dark vertical line on the plot. To account for possible non-linear associations between the propensity to negotiate one’s salary and the macro-level variable, we recoded the Gender Equality Index into four categories, one for each quartile.
Logistic regression models are used to test our hypotheses. The first hypothesis is tested by including the dummy for gender (women vs men as reference) while for the second hypothesis, we include the interaction between gender and the four-category variable capturing levels of societal gender equality.
The models also include certain socio-economic background variables, namely: “Living with/without children (dummy)”, “Perceived importance of your voice in your own country”, “Life satisfaction”, and “Economic difficulties”. As regards the second control variable, the respondents were asked to what extent they agreed or disagreed with the following statement: “My voice counts in (our country)”. For the next control variable, the question was “Are you very satisfied, fairly satisfied, not very satisfied, or not at all satisfied with the life you lead?”. Finally, “Economic difficulties” measures if and how often the respondents were unable to pay their bills during the last year. The coefficients of the control variables are not shown but are available upon request.
The results from the logistic regression in Model 1 indicate that women are less inclined to negotiate their salaries, as expected by H1. Indeed, the coefficient is negative and statistically significant. The predicted probability for men is 0.59, while for women it is 0.48, with a difference of 10.4 percentage points in salary negotiation.
The main effect of the Gender Equality Index is not statistically significant, with the exception of the second quartile. However, the magnitude of the coefficient is small.
Interestingly, when we include the interaction between gender and the gender inequality measure, gender loses statistical power.
We also note that the coefficients for the interaction terms are now negative and statistically significant. This indicates that there is an interaction effect between the two variables of interest and that the gender gap in salary negotiation increases at higher levels of gender equality, against our second hypothesis.
|Model 1||Model 2|
|Woman||-.442 (.041)***||-.150 (.080)|
|Gender Equality Index Quartiles|
|2nd quartile||-.130 (.059)**||-.014 (.086)|
|3rd quartile||.028 (.057)||-.223 (.086)**|
|4th quartile||.042 (.067)||-.376 (.095)***|
|Women × Gender Equality Index Quartiles|
|Woman × 2nd quartile||-.267 (.117)*|
|Woman × 3rd quartile||-.356 (.114)**|
|Woman × 4th quartile||-.622 (.122)***|
|Constant||-.415 (.162)***||-.575 (.168) ***|
To grasp a better understanding of this interaction, Figure 2 plots the predicted probabilities, separately for women and men, of being willing to negotiate their salary at different levels of GEI, while controlling also for all the other variables included in the model.
From the graph, we can observe that, for men, the propensity to salary negotiation keeps increasing as GEI increases; this is not true for women, whose comfort in negotiating their salary decreases from the first to the last quartile of GEI.
Thus, as societal gender equality, the gap between men and women increases too: from a 3.6 percentage point difference in the first quartile to 18.2 in the fourth quartile.
Summary and conclusion
The aim of this short paper was to examine the association between gender and the propensity to negotiate one’s salary.
Overall, as shown in previous studies and in line with our first hypothesis, women are less inclined to negotiate their salaries compared to men.
However, contrary to what we hypothesized in H2, the gender gap in salary negotiation is larger in countries with higher scores of societal gender equality, measured via the Gender Equality Index.
These results suggest that a context characterized by low gender equality affects people in the same way regardless of gender. In contrast, better conditions, ceteris paribus, emphasize differences between genders, benefitting men and disadvantaging women.
Further studies could investigate these results in a deeper way, focusing on the reasons behind these differences, especially in countries with good labor market conditions.
Moreover, additional analyses should account for whether other macro-macro level indicators, such as inequality, wealth, or human development also account for the observed pattern.
To conclude, considering our results, investigating more thoroughly socio-economic conditions and their effect on salary negotiation would be an interesting starting point in future research.
- Babcock, L., & Laschever, S. (2003). Women don’t ask: negotiation and the gender divide. Princeton, NJ: Princeton University Press.
- Johnson, J. (2016). Gender Differences in Negotiation: Implications for Salary Negotiations, UCLA WOMEN’s L.J., 23, 131-151.
- Lauterbach, K. E., & Weiner, B. J. (1996). Dynamics of upward influence. Leadership Quarterly, 7, 87–107.
- Riley Bowles, H., Babcock, L., & Lai, L. (2007). Social Incentives for Gender Differences in the Propensity to Initiate Negotiations: Sometimes It Does Hurt to Ask, ORGANIZATIONAL BEHAV. AND Hum. DECISION PROCESSES, 103, 84-103.
Foto di Gabrielle Henderson su Unsplash