Partisanship, protection, and punishment: how governments affect the distributional consequences of International Monetary Fund programs

Women's leadership and the gendered consequences of austerity in the public sector: Evidence from IMF programs

“…We focus on public‐sector employment outcomes for several reasons. First, contrary to employment outcomes in the private sector which are mediated through firm‐level decisions (Reinsberg & Abouharb, 2022), governments can directly affect employment outcomes in the public sector. Second, public sector employees are frequently the first ones to lose their jobs or swallow deep cuts in salary and employment benefits (Berik, 2016; Gunaydin, 2018; Reinsberg, Kentikelenis, et al., 2019; Rickard & Caraway, 2019).…”

“…In the appendix, we plot the number of active IMF programs per year (FigureA1) and the average number of binding conditions across programs (FigureA2).3 ILO employment statistics, various years, drawn from https://www.ilo.org/shinyapps/bulkexplorer50/?lang= en&segment=indicator&id=EMP_TEMP_SEX_ECO_DSB_NB_A (accessed 1 August 2022).4 ILO employment statistics, various years, see earlier footnote.5 https://data.oecd.org/earnwage/employee-compensation-by-activity.htm#indicator-chart.6 Research supports the notion that increasing the number of women within an organization and women's leadership have the power to weaken stereotypes about gender roles(Adams & Funk, 2012;Beaman et al, 2009;Dahl et al, 2021;Kunze & Miller, 2017).7 An alternative instrument using the interaction between the IMF liquidity ratio and the probability of an IMF program was too weak, likely due to our restriction to developing countries and a shorter period(Lang, 2020).8 Empirically, this bias is less relevant, because we find that cabinets with women are more likely under an IMF program in our sample. If at all, this would induce bias against our findings.9 Random-effects multi-level regression is not advisable because the country-level treatment of an IMF program is not randomly assigned, which would produce biased estimates(Daoud et al, 2017;Papke & Wooldridge, 2022;Reinsberg & Abouharb, 2022).10 For instance,Laird (2017), analyzing administrative data from the U.S., finds that racial minority women were disproportionally affected by public-sector downsizings during the Great Recession.…”

“…9 9 Random‐effects multi‐level regression is not advisable because the country‐level treatment of an IMF program is not randomly assigned, which would produce biased estimates (Daoud et al., 2017; Papke & Wooldridge, 2022; Reinsberg & Abouharb, 2022). …”