A study by academics in Sweden combines anonymous responses from a Swedish government survey with employment data to study sexual harassment in the workplace. They found that about 13 percent of women and 4 percent of men reported being sexually harassed in the previous 12 months. The more male-dominated the workplace, the more likely women were to have been harassed. Conversely, the more women dominated the workplace, the more likely male workers were to have been harassed. The highest rates for women were in manufacturing jobs, while the highest rates for men were in service and sales jobs. Power dynamics seem to play a role here: both men and women reported more sexual harassment when their supervisor was of the opposite sex. Women who self-reported harassment were more likely to switch to new workplaces with more female colleagues, where pay tends to be lower. Interestingly, male victims did not migrate to more male environments where pay tends to be higher. The researchers conclude that sexual harassment contributes to the perpetuation of gender segregation in parts of the labor market by preventing individuals from working in places where they belong to the gender minority. What are the consequences for the perpetrators? To that question, a new working paper offers some tentative answers. It links information on every police report in Finland between 2006 and 2019 with administrative records on employment, income and demographic characteristics. Abi Adams-Prassl, professor of economics at Oxford, and her co-authors identified more than 5,000 incidents of violence between co-workers who shared their workplace. The vast majority of the perpetrators were men, while the victims were evenly split between men and women. After a violent episode, both perpetrators and victims experience a drop in income and employment. But there was a strikingly different pattern by victim gender. After a male-to-male incident, employment rates dropped by an average of 10.6 percentage points for the perpetrator and 4.2 percentage points for the victim over the next five years. However, after an incident between men and women, employment rates decreased by only 5.2 percentage points for the perpetrator and 8.4 percentage points for the victim. (As for the incidents where women were the perpetrators, Adams-Prassl said there were too few of them to analyze.) Like the Swedish study, a power imbalance helps explain what’s going on. The Finnish study finds that when perpetrators are older in the workplace, there are fewer career consequences. Researchers find that victims of male-female violence are relatively young and low-income compared to their perpetrators. This is not the case for male-on-male violence, which tends to occur between relatives who are equal in age and income. The impact of these incidents spills over into the rest of the workplace. After a male-to-female incident, the gender composition of the workforce becomes significantly more male (there is no change after a male-to-male incident). This is due both to more women leaving and fewer women joining. However, it seems there are ways to stop these leaks from happening. Workplaces with more female managers (defined as the higher-than-average share of women in the top 20 percent of earners) do not experience a decline in the share of women in the workplace after an incident. They do not better protect victims’ careers, but seem to punish perpetrators more effectively, based on the fact that they are more likely to suffer a drop in employment. The authors summarize: “Women managers do one important thing differently: perpetrators fire.” A lot of this may sound intuitive, but having the hard data matters. The bad news is that co-worker harassment and violence can have serious and disproportionate effects that extend far beyond the individuals involved. The good news is that we are at least beginning to calculate the scale of the cost and who is paying it. [email protected]
