Tyler Cowen points me today to a new study that investigates whether the gender wage gap is affected by family leave laws. The basic strategy is to look at the gender gap by state and compare wages before and after family leave acts were passed. That's a sensible approach.
Long story short, the authors conclude that family leave laws do indeed affect the wage gap. Here's their chart for the gender wage gap on a national basis. To avoid racial confoundments, they compare white women to white men.
The vertical red line shows the passage of the Family and Medical Leave Act under Bill Clinton. Sure enough, women were quickly closing the gap until then, but after that their progress slowed down dramatically. This is a little weird, though, because I've never seen a chart like this. Here's the same thing using ordinary old wage income available from the BLS and the Census Bureau:
The BLS counts only full-time workers while the Census counts everyone, which is why the numbers are different. However, they both show similar progress toward closing the wage gap and they both show that progress staying about the same until the Great Recession in 2009. Nothing at all happens in 1993.
But let's assume the study authors did their arithmetic more carefully than me and something really did happen to the wage gap in 1993. They acknowledge that it's difficult to blame FMLA alone for this, since both welfare reform and an EITC increase were passed around the same time. However, after using up the entire Greek alphabet they conclude that it was indeed FMLA at work. In fact, FMLA accounts for virtually the entire effect.
I'm already a little skeptical for two reasons. First, the authors use a tremendous amount of modeling to get their results. Second, it's too perfect: FMLA passed and almost instantly progress on the wage gap slowed down substantially. That's a little hard to believe.
But let's continue anyway. Just exactly what was FMLA responsible for?
In the decade prior to the passage of the family-leave policy, wages for white men are stable. After the passage of a family-leave policy, wages for white men grow steadily to levels that are more than $1 per hour higher.
By contrast, [] prior to the implementation of a leave policy, white women’s wages are increasing steadily to a level that is nearly $1 per hour higher than one decade prior. After the policy, white women’s wages continue to grow at nearly the same annualized rate compared to before the policy.
In short: The growth of women's wages were unaffected but the growth of men's wages accelerated.
I just don't know about this. Aside from my other issues, I can't think of a mechanism for this. Why would men's wages suddenly accelerate after passage of a family leave act?
The authors don't try to address this, which is fine. If they don't know, they don't know. Still, you'd really like to have some idea of how this could work, especially when the results are so head scratching. After all, most European countries have better family leave policies than us and they have lower gender wage gaps:
Maybe I'm missing something here, but if I am I don't know what it is. For the time being, put me down as agnostic about this.
More smoke thrown up against the narrative that workers, be the collared blue or white, are insisting on their rights as humans, laborers, and family members.
Start making more money weekly. This is valuable part time work for everyone. The best part ,work from the comfort of your house and get paid from $10k-$20k each week . Start today and have your first cash at the end of this week.
Visit this article for more details.. http://incomebyus.blogspot.com/
Pretty sure it was Kurt Cobain dying. Or maybe the players' strike. Or something else I disapprove of that caused the bad thing to happen.
+1
Why would men's wages suddenly accelerate after passage of a family leave act?
Well, one possible causative mechanism might be: A) if mothers tend to take advantage of family leave more than fathers, and B) the policy in question tends to raise costs for employers*, then C) employers might then exhibit a preference for male employees. And so the latter's wages rise relative to female employees.
Shouldn't women's wages then fall as there is less demand?
Shouldn't women's wages then fall as there is less demand?
Not necessarily. We'd also need to know about supply. And we moreover don't know there's "less demand." There may be increasing demand—but it's not increasing as rapidly as the demand fo male workers.
I think you're describing how it was modeled. But does that fit the data? The authors would have to be arguing that other factors were really helping to close the gap, but FMLA held them back to give us the data we see.
No, I haven't worked through the original paper or more importantly, run their models.
The FMLA could mean more women take time off to raise/deal with family, lowering salary. It could also mean that those women can come back into the workforce and restart where they left off, and not back at the bottom, minimizing wage loss due to their absence.
Perhaps FMLA was responsible for the US economic boom and resulting pay increases that occurred in the mid to late 1990s?
Or, given other highly selective analysis we can guesstimate that the Reagan tax cuts and union busting which caused the economic boom a decade later were also responsible for the gender pay gap widening.
we can believe any priors that we want
+1
I think maybe one ought to break down the wage gap by sectors. Perhaps there's a different story to tell about structural changes to employment due to globalization and specifically NAFTA and opening up CN-US trade.
I bet this would be enlightening.
My understanding is that worker productivity in the EU is substantially less than the US and some have attributed this to very generous leave policies (maternity, vacation, sick) for men and women in the EU. In the US we have unfurled maternity leave with women in mind, but having critical personnel (e.g. fiscal, HR) potentially leaving for extended periods of maternity leave requires redundancy in the system. This means the redundant pieces need to be paid less or you can hire someone who will not take maternity and pay them more. Once we have equal maternity for men and women, I suspect this pay gap will disappear.
Several cautions on productivity comparisons: 1) Differences between the US and Northern Europe are not very large, and Northern European countries generally have the most comprehensive ‘social safety nets’; 2) There are multiple ways of measuring productivity, and the one that economists use (GDP/hour worked) is about the worst, as it is distorted by factors that have nothing to do with the efficiency of delivering real goods and services. The countries with the highest GDP/hour worked are oil/gas states, tax havens (Ireland) and financial centers (i.e. banks with parliaments like Luxembourg).
More than 90% of employers are smaller than 50 employees so lots of workers have no access to leave. I wonder if women seek jobs at the few employers that offer leave, thereby lowering their wages to get the better benefits?
Smaller firms are not legally required to offer leave, but many still do so, albeit maybe not for the same period under the FMLA. I've worked at such firms so I would not make the claim that because the leave is not legally required it does not happen.
Around 20% of workers have access to paid family leave.
I'm not disputing that. The FMLA does not specify PAID leave, but it's inaccurate to claim that people at >50 employee firms have no access to leave.
IThe gap as a percent difference in gender pay disparaties can be closing even as the absolute difference is increasing. An implausible example would be women going from one to ninety-eight dollars an hour, men from two to onehundred over the coure of a single year. So which is it when it's reported that 'the gap is closing'? I know it's a reach.
And how are "wages" defined? If their all developers, then their net income is always zero--which I have from good authority (a Republican congressman).
With the growth in income inequality, I'd thought the median, not average, value would work best for this type of study.
Is this the same Tyler Cower who two weeks ago didn't accept the zillions of charts showing correlations in various countries between blood lead levels and crime stats?
Tyler Cowen, of course
A big red flag for this study is that the authors had wage data from 1975 to 2015, but instead of plotting the actual data (like Kevin did in his graph), they decided to create a model, and then used the model to draw their conclusions instead of the actual data. Models are good tool when trying to predict the future, or when you have datasets with a lot of data gaps, but neither case applies here.
Seems to me like a not so clever sleigh of hand to arrive at a desired conclusion despite the data available.
That’s why economists are so often the butt of the joking remark, “It works in practice, but will it work in theory?”
Intuitively Cowen's finding feel much to extreme. In contrast, I would be very surprised if the FMLA didn't have a consequential impact: an axiom of economics, you make something more expensive and you typically get less of it.
oops much too extreme
Many "institutions" consider these benefits as salary. Maybe if men got family leave, their salaries would go down....
What asks explanation is the above-trend growth of men’s wages“The technology sector has a dynamic history of expansion and contraction. Its first high-growth period lasted from 1990 to 2000, a time traditionally thought of as the “dot-com boom” or the “tech bubble.” National employment in technology sector industries shot up by 36 percent over the period. (See Figure 1.) Average weekly wages for technology sector workers doubled, rising by 102 percent over the 10-year period.”
https://www.stlouisfed.org/publications/regional-economist/second-quarter-2017/growth-in-tech-sector-returns-to-glory-days-of-the-1990s
So, very high wage growth in a field that disproportionately employs males, plus a composition effect from employment growth in that field?
An artifact of lower paid men dropping out of the workforce?
I've looked at studies dine in the Nordic countries where you have the most equality, and what happens is that all things equal the person farther in their career stays at work rather than equal leave because it makes more money. Because in most marriages that will be a man either due to patriarchy or because the male is older, rational choice reinforces any gaps.
"Maybe I'm missing something here"
You're missing that cowen is a font of stupid shit, both his own stupid shit and stupid shit he promotes from others.
I've been wanting to make a comment all day, but need to keep it under 10k words. golack made a very important comment "With the growth in income inequality, I'd thought the median, not average, value would work best for this type of study." This is extraordinarily important. Per the Congressional Research Service, the real median wages of men declined by 3% between 1979 and 2019. https://sgp.fas.org/crs/misc/R45090.pdf
My recollection is that for most of the past 50 yrs, real median wages for white men have been less than they were in 1973. So women have reduced their wage gap with white men a fair amount, and black men have reduced theirs a bit. But the target they are shooting for hasn't moved up at all. It used to be that wages for everybody went up as productivity went up, but that relationship died in the 1970s. Most women would be better off financially today if wages had stayed tied to productivity, even with the more inequitable pay gaps of the past. (See this: https://time.com/5888024/50-trillion-income-inequality-america/)
The final thing to say is that concentrating on wages ignores the fact that the people at the top take a trivial percentage of their income from wages. If you look at household income from all sources you get a better grasp of the inequity. From the early 1940s to the mid 1970s, the bottom 90% of households got about 63% of all personal income in the US. This went up in booms and down in recessions, but the general trend was up. If you plotted the curve in 1970, you'd expect 65% to be the average today. Instead, the last figure I saw was 49% going to the bottom 90% of households. That doesn't mean the 91st or 95th percentile are getting their fair share either. Almost all of this shift in income has gone to the top half or top tenth of the top one percent.
Put it another way, the average household in the bottom 90% of households would have about $22k more in annual income today if the rules of the game hadn't been changed to make the rich richer.
It's coincidence. White men's wages went up in the 90s because of the boom in tech, a massively male-dominated industry. FMLA had nothing to do with it.