To address gender inequality, it is necessary to first identify its existence and size and then provide appropriate incentives for companies to make serious efforts to reduce it.
In July 2022, the Japanese government mandated companies with a workforce of more than 300 employees to disclose information on their “gender wage difference.” Under its “policy to prioritize women empowerment and gender equality,” those companies are required to disclose the ratio of the wage level for women to that of men for both regular and non-regular employees. However, if the average wage levels for men and women are simply compared while ignoring other variable factors, the results would reflect differences not only by gender but also by age, educational background, and years of service.
Let us consider this problem by looking at concrete examples. In the case of chemical manufacturer Company A, a large proportion of employees in the production division are men with only a senior high school diploma, while the research and development division has many female workers with a master’s or doctor’s degree. In other words, a simple comparison in this case would be between the averages for a female employee group represented mostly by high academic achievers and a male employee group that included many low academic achievers. Under this comparison, gender wage inequality within the same type of job would be obscured by the difference in gender distribution across job types.
On the other hand, in the case of food manufacturer Company B, the majority of employees in management positions are men, while most women are engaged in junior clerical jobs. If the group of female employees, most of whom are relatively young and who are mostly concentrated in lower brackets in terms of both educational background and years of service, is compared with the group of male employees, who vary widely in terms of those attributes, gender wage inequality would be significantly exaggerated by differences due to other attributes.
The greatest problem of all is that a simple comparison of the average wages for male and female employees does not deliver meaningful results across companies and business sectors. As the wage difference calculated in this way would not accurately reflect efforts to improve equality, pressure from the labor market and/or capital market will not induce firms to take desirable actions even when disclosure of gender wage difference is required.
Switzerland has been ahead of other countries in increasing the visibility of gender wage differences. In 2006, the country’s federal government started promoting the use of a web tool that uses regression analysis to calculate gender wage differences while controlling for basic attributes that may affect wages, such as age and educational background. Companies with a workforce of 50 or more employees that have been found to have a gender wage difference of 5% or higher are subjected to sanctions, such as limited participation in government procurement programs and limited access to subsidies. It was reported that since the introduction of this regulatory measure, the overall gender wage difference has narrowed 3.5 percentage points. In Japan, a research team to which I belong has also developed and started providing a similar software tool to companies.
◆◆◆
Many economic studies have focused attention on the effects of wage premiums paid for long working hours and the gender division of labor within the household. A typical empirical observation is that the longer employees work and the more frequently employees work overtime at night or on holidays as needed, the higher the hourly wages are. The claim that the presence of wage premiums paid for long working hours is a major factor that creates a persistent gender wage inequality is supported by several studies (Goldin 2014).
While it is relatively easier for men to engage in jobs that require long working hours, women face greater restrictions on the types of job they can choose and how they work due to the heavy burden of household chores placed on them, resulting in wage inequality. In Japanese companies, where jobs are less standardized than in the United States and Europe, workers are less substitutable with each other and know-how is not systematically shared, but accumulated in a fewer number of individual workers, making it easier for the wage premium for long work hours to be larger.
The gender wage inequality is particularly conspicuous for women after giving birth. It is because many women give up their career or make a career shift to the “mommy track,” which makes it easier for female employees to achieve work-life-balance, but offers dim prospects for career advancement. The income loss associated with giving birth and childrearing, known as the “child penalty,” is particularly large in Japan.
To correct the disproportionately heavy burden borne by women in terms of household chores and childrearing, it is necessary to promote a change in the ingrained gender division of labor, which obliges women to concentrate on attending to household duties while men work outside the home. The government should develop an environment that makes it easy for employees of the childrearing generation to take child-care leave. To that end, it is essential to introduce obligations for disclosure of the percentage of male employees who have taken child-care leave and the duration of the child-care leave taken, and to require employers to identify the difference in the rate of promotion between employees who have taken child-care leave and employees who have not for both men and women. Younger generations of both male and female employees tend to attach greater importance to work-life balance, and as a result, companies that fail to take action to accommodate that need are expected to have an increasingly smaller pool of potential job applicants.
Reducing work processes that depend too much on the skill of individual employees allows for flexible ways of working. It is necessary to standardize work processes and skills, create appropriate task manuals, visualize task progress through IT and introduce more teamwork-based processes. In other words, work processes should be improved so that employees can substitute for each other when necessary. Increasing the visibility of work processes facilitates efficiency improvement and is also consistent with efforts made by many companies to shift to a job-oriented employment arrangement and to systematize skills development.
In recent years, studies in social sciences have shown that social prejudices concerning gender differences and the ensuing discriminatory or biased behavior have contributed to the creation of gender inequality. Two major sources of gender biases are “statistical discrimination” and “gender stereotyping.”
Statistical discrimination refers to forming certain presumptions about the characteristics of individual persons based on the average characteristics of attribute groups to which those individuals belong and acting toward them based on those presumptions. One example is for managers to presume that women are more likely than men to quit a job and, under this presumption, give precedence to male employees over female ones in providing growth opportunities to attend training programs or work on challenging projects that may lead to a successful career.
Meanwhile, gender stereotyping refers to forming wrong assumptions based on gender, such as that men are fit to be in leadership positions while women are better suited to supporting roles. Many people unconsciously expect men and women to behave in ways that match such gender stereotypes. Individuals who do not resemble the gender stereotypes, such as “women with strong leadership skills,” for example, tend to be seen in an unfavorable light and be held in low regard by society at large.
◆◆◆
To make earnest efforts to resolve gender inequality, it is essential to promote data usage and to understand at which stage of employees’ career development gender inequality tends to widen. Among the main issues are: whether the same recruitment criteria are applied to men and women; how severe the child penalty is; what the level of difference in gender mix by type of job and by function is; at which career stages do gender differences in promotion emerge; and what level of gender difference exists in terms of the duration of working hours, the amount of tasks assigned, the difficulty of annual goals, the participation in training programs, and the range of experience offered through personnel transfers including those across markets and borders.
To develop effective measures to encourage efforts to resolve gender inequality, it is essential, first and foremost, to identify important indicators and monitor developments. Data thus collected will be useful for verifying the effects of those efforts.
The figure below shows the distributions of performance and competency evaluation scores by gender in one company. While there is little gender difference in performance evaluation, there is a significant difference in competency evaluation.
In cases where a certain level of competency evaluation score is required for assignment to a management post, the sort of difference in evaluation scores observed in this case would result in a significant gender difference in the pool of potential candidates for management posts. The main factor that created the significant gender difference in the competency evaluation at this company was the difference in evaluation of leadership skills. It is necessary to scrutinize the case by holding interviews with employees and identify necessary measures weighing various possibilities, such as whether women are inferior in leadership skills, whether women cannot exercise leadership due to the fear of behaving in ways that do not conform to gender stereotypes, and whether evaluating superiors are compelled by a masculine image of leadership.
It is desirable to develop a system to collect data and monitor necessary indicators and to formulate an action plan. The most important thing to do is for companies to clarify institutional responsibilities and for management teams to commit themselves to correcting inequality where it exists. They need to understand that disclosing both specific numerical targets and the specifics of the measures that are planned and taken is effective in increasing their reputations in both the labor and capital markets.
* Translated by RIETI.
December 16, 2022 Nihon Keizai Shimbun
