This paper uses a 2016–2024 individual-level panel to test whether COVID-19 changed public attitudes toward women’s participation in the workplace. Using fixed-effects and fixed-effects ordered-logit models, the authors find that (1) attitudes became more positive after COVID-19, (2) older cohorts shifted from more negative to more positive views post-pandemic, and (3) married men and certain worker-status groups exhibited distinct changes; these results are robust across specifications
速读内容
Key trend: Men vs Women — rising support for women’s work after COVID-19
Both genders show increased mean support for women’s active work after the pandemic; women’s mean rises more sharply and by 2021–2024 women exceed men in average support .
The authors interpret this as a COVID-driven recognition of women’s role in essential services and labor shortages in aging Japan .
Distribution shift: choice-category histograms before vs after COVID-19
Before COVID-19 the modal response was “3=neutral” for both sexes; after COVID-19 the modal response for women shifts to “4=Agree” (≈40%), while men increase in 4 and 5 but remain more concentrated at 3 .
This indicates both a central tendency shift and a polarization toward stronger agreement for women in particular .
Sample summary: key variables (means and s.d.)
Variable
Definition (short)
Mean
s.d.
View Woman Work
5 (Strongly agree) − 1 (Strongly disagree)
3.38
1.28
Covid
Dummy (2021/2023/2024 = 1; 2016/2017 = 0)
0.44
0.49
Age5064
Age between 50 and 64
0.43
0.49
Age65
Age ≥ 65
0.08
0.37
Married
Married = 1
0.63
0.48
Full-worker
Full-time worker = 1
0.56
0.49
Sample used for main FEOL reported: 13,265 observations; full FE sample ~22,649 observations after controls .
Econometric findings: FE and FEOL — core results and magnitudes
COVID-19 dummy (Covid) is positive and highly significant in FE models: public attitudes toward women’s active work increased after the pandemic .
Interaction terms Covid×Age5064 and Covid×Age65 are positive and significant: older groups (50–64 and especially 65+) exhibit larger positive shifts after COVID-19, with Covid×Age65 showing the largest coefficient and strongest significance .
Heterogeneity by gender: subsample FE results show the pattern holds for both women and men, though magnitude and some controls differ (e.g., Covid×Married is positive for men but negative for women in FE subsamples) .
FEOL marginal effects (probability of choosing “strongly agree”): Covid×Age65 marginal effects ≈ +14.0 pp (full sample), +18.6 pp (women), +14.3 pp (men); Covid×Age5064 marginal effects ≈ +7.5 pp (full), +10.6 pp (women), +6.7 pp (men); older respondents show larger probability increases of strongest support post-COVID .
Mechanisms and interpretation
Authors argue the shift reflects increased visibility/valuation of essential health and care work during the pandemic and labor shortages in an aging society, prompting especially older people to prefer greater female labor participation .
Married men’s increased support is consistent with household income-risk considerations (wives as potential breadwinners) during exogenous shocks .
Robustness and limitations
Results hold in FE and FE ordered-logit specifications; marginal effects reported to quantify substantive impact .
Limitations: unbalanced panel, inability to fully separate aging-trend effects from COVID effects, and generalizability to other countries needs further study .
报告标题:“Has the COVID-19 Pandemic Altered the Traditional View about Women’s Active Work?”,作者为 Eiji Yamamura(Seinan Gakuin University)与 Fumio Ohtake(Osaka University)。报告声明无利益冲突,数据/代码可向通讯作者请求获取。
主题/研究对象:研究新冠疫情(COVID-19)是否改变了公众对“女性积极参与职场(women’s active work)”的看法,数据为2016–2024的个人层面面板调查(网络问卷),分析采用固定效应与固定效应有序Logit(FEOL)等方法。
数据来源:Nikkei Research Company实施的网络问卷,2016、2017、(2018缺少关键问题)、2021、2023、2024 共6次点位的面板追踪,其中2018波未包含关键问题因此被排除,导致样本为2016、2017、2021、2023、2024这些时点的混合(不平衡面板)。
关键变量定义:被解释变量为对陈述“The government should create a society where women can fully demonstrate their abilities and play an active role in the workplace.” 的5点Likert评分(1-5),Covid为2021/2023/2024的哑变量,Age5064与Age65为年龄分组哑变量,Full-worker为是否为全职员工等控制变量;样本描述及均值见Table 1(样本观察数、变量均值和sd等)。
模型:作者设定面板固定效应模型(以个体固定效应 ki 控制不可观测时间不变特征),主要回归形式包含Covid与年龄分组的交互项(Covid×Age5064、Covid×Age65)以及Covid×Married、Covid×Full-worker等控制交互项。因被解释变量为有序变量,作者同时使用FE有序Logit(FEOL)作为稳健性检验。