Research & Action Spring/Summer 2015

Research & Action Report, Spring/Summer 2015

Yes, in a number of cases. Recently we did an intervention in Finnish high schools where we provided information to graduating students about employment opportunities and salaries related to the fields of further education they were considering, and then checked for resulting changes in their choices. The results showed very clear gender differences emerging in their choices and in the factors, including the information we had given them, that went into these choices. I think those results will be very relevant here as well; we can see very similar choices made by U.S. students.

I’ve also been working on labor market polarization. In Finland we seem to be losing a lot of mid-level jobs that previously provided nice middle-class incomes, while more and more jobs are being created at the lower end of the job spectrum. This is also happening in the U.S., but here we don’t always have very good data to study that phenomenon. So we can use some Finnish data to try to understand, for example, whether this is happening because new firms are entering the market while some firms are exiting, or whether existing firms are changing their job structure and out-sourcing some of their jobs. It will be very informative to see what the actual dynamic process is behind the job transformation.

And then of course there’s the parental leave work that we have initiated in Finland and that I’m working on now in the U.S. There are many lessons that can be taken from the Finnish experience.

The Family and Medical Leave Act [FMLA] of 1993 entitles many U.S. employees to take up to 12 weeks of job-protected, unpaid leave for medical or family reasons. What emerged from our research was a very uneven picture of who actually has parental leave available through their work, who is able to take it if it is available, and how long such women will stay on leave. It turns out that mostly fairly high income women are actually able to take any advantage of it.

If you look at women in, say, four buckets according to their family income, then the upper two buckets are doing fine and the lower two buckets are not doing well at all in terms of the availability of leave and the ability to use it. A lot of lower-income women work multiple jobs and don’t have enough hours in any given job, or their periods of employment have been too short to make them eligible for the leave. Even if they are eligible, this is unpaid leave, so many women cannot take advantage of it because they’re living paycheck-to-paycheck without any financial cushion to fall back on.

So what are needed are more paid-type leaves. A few states have already instituted a version of paid leave; and a lot of firms across the country actually do have paid leave available, although not necessarily for all their employees.

The goal of our project is to capture data from inside about 500 firms that offer paid parental leave and to follow all their employees over time, to learn what happens for the employees and for the firm. That’s what makes this project different. Most people studying parental leave or related issues use data from various U.S. surveys, but most of these are cross-sectional surveys, where you survey one person, let’s say in 2015, but the same person won’t be surveyed again in 2016. So you just get snapshots of people, and they’re in various firms.

But I want to learn what happens over time to an employee who starts with a firm, and maybe two years later has a baby. In a family-friendly firm, is she likely to return after the parental leave? Is she more likely to stay with the firm over a long period of time? Is she having steady salary growth, and any sort of overall career progression? Of course, the parent could be a father, but it’s mostly female employees who use these leaves.

I also want to study the effect of paid parental leave on firms themselves. As I’ve said, many firms actually do have paid leave now; so it must be a good idea for the firms or otherwise they would not be doing this. Employers aren’t doing it out of the goodness of their hearts, they’re doing it to attract and keep qualified female workers. Recruiting is expensive, it’s expensive to lose women who have very specific human capital that they’ve acquired through their careers, so we might assume the firms are finding it profitable to offer these leaves. But we want to get a better understanding of how they are actually doing.

I was in Washington, D.C. before Christmas and was invited to visit, just socially, the Department of Labor. I got to talk to the Secretary of Labor for a little while, and their chief economist as well; and paid parental leave is very high on their list of concerns. They were highly supportive, they were interested in my research, but they were not very hopeful that any kind of federal legislation could be passed, believing that it would be more a state-by-state effort. California, New Jersey, and Rhode Island now have paid-leave policies, and I’m sure more and more states are going to follow suit.

And the more we show evidence that this is good for firms, the more likely it is that the policy will spread. If you have a ton of research showing that this is good for the children and good for the mothers, but you don’t have research showing that it’s actually not that detrimental to the firms, it’s not likely to make much headway. Understandably, such a policy is hard for small firms. But once we’re got good data from firms, we can start the discussion. If it is detrimental to some firms, how can we structure it in such a way that it doesn’t impact them too negatively?

We won’t begin with 500 firms! We’re starting a pilot part of the project working with about 30 firms from the Working Mother Magazine’s list of “100 Best Companies” to work for, calling them the family-friendly firms. First we need to confirm what their family policies are. We’ll learn what we can from available documents, and then try to have a two- to five-minute phone interview with each head of Human Resources [H.R.] to confirm that this is in fact the way their policies operate. For help with introductions to these people, we may even use the Wellesley College network; I’ve found that there’s someone from Wellesley at every large firm. Wellesley women are everywhere—even in the Department of Labor! We’ll tell each H.R. person, “We’re doing research and your firm has been deemed one of the best places for women to work. Could you confirm for us that the company policy is such-and-such?” Ideally we will ask only simple yes or no type questions, and obviously we won’t use any of the firms’ names in our research.

Then I’ll go into the confidential, very detailed firm-worker-level data in the U.S. Census Bureau’s Research Data Center system, using the data center right here in Cambridge [MA]. I can find my family-friendly firms and comparison firms in there. The data in that system is just unbelievable. You can follow every single person over time—not by name, of course, the social security numbers are all encoded, so I never see any of them. And you can link people across data sets, so I can have Decennial Census data that can be linked into the firms’ employee-level data sets, and so on. It’s absolutely perfect data. There’s unlimited potential in those data for following firms and people over time and doing everything from creating useful graphs to using state-of-the-art econometrics.

There are two reasons. First, most people don’t have access to it. Theoretically, it’s accessible to any researcher who wants to undertake research using those data. But you have to have “special sworn status” with the Census Bureau, and then pass a full background check. You must also have a research proposal that is approved by the Census Bureau and the IRS. I think it’s a two-year-long process until you finally get access, and then there’s a big learning curve involved. You need specialized programming skills in order to operate in that environment.

Another big challenge is that this is exploratory, trial-and-error work. Those data sets are gigantic. Some of my regressions have maybe 80 million people in them! That’s exciting, but until I start really doing something serious, I can’t know exactly what’s going work. Maybe we can learn something totally different, or maybe we can confirm that what we thought is true is actually true, using data that are much more extensive and detailed. Luckily the NSF [National Science Foundation] and other foundations have understood this about what I am proposing. And the project is going to teach every other researcher something as well.

There’s a lot to do. Fortunately, I have lots of experts working as my co-authors with me. And if I can’t get some things done, better people can get it done with me.

Actually, some people have said that about me, that they trust I can make things happen. The main thing is to convince the people who have the funding. Once you’ve shown one organization that you can get things done, it gets easier to convince other people as well.

The subject was that research I mentioned earlier in which we asked students in their last year of 60 Finnish high schools about what further education and degree programs they were anticipating, and why. Were they considering the future employment prospects and salary probabilities of their choices? What were their attitudes toward risk? We learned up front that a lot of girls are going into education, humanities, or some type of social science, whereas boys are more inclined to say engineering, business education, law. We also learned that girls were more likely to than boys to have over-predicted what their future salaries would be.

Then we gave them information about their prospective employment opportunities and average salaries, and checked for changes in their opinions. We found that for both boys and girls employment prospects and salary matter; it’s not true that girls don’t care about those things. But girls actually care more about their interest in their future fields than the boys do. If boys from low-income geographical areas were negatively surprised about the salary probabilities in their future fields, they were more likely to switch to higher-paying degree programs. But girls didn’t budge a little bit! They hadn’t gone to university yet, they were still six or seven years out of the labor market, and already there was your gender pay gap!

By the time boys and girls are high school seniors, they’re fairly set in their ways. If we want to make a difference, we need to do something much earlier. For example, there’s a big push to get girls into STEM [science, technology, engineering, and mathematics], but it will need to happen very, very early.

We have at least two distinct streams of research. The first is related to immigrant entrepreneurs. There are a lot of policy efforts at the federal level, White House level, and state and city levels, to attract skilled immigrant entrepreneurs into the U.S. to found firms. Different visa categories have been created—for example, there’s a “million dollar green card.” Somebody who invests a million dollars in a firm in the U.S. gets an automatic green card. But we don’t really know anything about the effects. Let’s figure out how many immigrant entrepreneurs there are, where they go, what kinds of firms they found, how many jobs they create—from the point of view of the funder, the NSF, job creation is actually the most interesting part. And what is the actual impact on the economy? Are those new jobs good jobs? Who gets them? How well do they pay? Are they all in Silicon Valley? So this work will focus on labor-market and regional growth impacts of immigrant entrepreneurs.

There will be a side piece focusing on the innovation of immigrant entrepreneurs. We’ve already found that high-skilled immigrants are responsible for a very large share of U.S. patents. So they’re innovative, but if they’ve started their own firms, how does that contribute to U.S. innovation?

In the second stream of research, we were approached by another foundation interested in the characteristics of new firms and founders in general. It may be hard to believe, given the current hype about start-ups, but in fact there’s been a steady decline in the start of new firms in the U.S. Part of the question is, what is that related to? Is it natural, since the population is aging? We want to research what has happened over the past 20 years—perhaps 1991 to 2011—in terms of characteristics of the founders. Is it the same kind of people always founding firms, or are the new founders different? How are their characteristics related to the firms’ ability to survive over the long run, to grow?

The foundation wants to measure not just things like the age and gender and education of founders, but also their innate thinking. Of course there’s no data set, no census anywhere in the world that would tell us anything like that. But luckily my co-author and I have a relationship with the huge Cambridge Innovation Center, which claims to house “more start-ups than anywhere else on the planet.” Its founder has very kindly agreed to allow us to go into the Center and actually work with the new founders who are currently based there. The Center has also surveyed every firm that has ever been there and has different kinds of data that they’ll allow us to use.

This work will be innovative and combines a lot of things that I haven’t seen done before, so it’ll be fun; and highly publishable and highly presentable work should come out of it.

Here’s one thing from the immigration work. A sort of anti-immigrant lobby argues that we’re already bringing in too many high-skilled immigrants who take jobs away from qualified U.S. workers. Then there’s a pro-immigration lobby of firms like Microsoft and IBM that say they can’t find enough qualified people. So we actually looked at this at the firm level—do we see high-skilled young immigrants replacing native employees when they arrive?

We found that for the most part, no, the firms seem to be using the immigration programs to grow, they’re not using them to replace anyone. There is a small effect in certain types of firms where it looks as though older native employees are more likely to leave, but we can’t tell whether these older workers are leaving first and then the firms are finding young immigrant workers to replace them, or whether it’s the other way around. Or whether the leaving is voluntary or involuntary.

But we don’t find that there’s any large scale replacing of educated U.S. workers by these young high-skilled immigrants. Immigration policy is one of those hot potatoes that nobody wants to touch, so our finding is helpful. It actually has been cited a fair bit in the press, so I know it’s having some impact in the debate

A lot of today’s economic work is related to women and their careers and to the family-friendly policies of firms, and/or to entrepreneurship. It’s interesting work, it’s timeconsuming work, and proximity matters. It would be very helpful to have someone local to bounce ideas off of, and to work with on grant proposals. Economics funding seems to be fairly robust; if you have innovative ideas, it’s not hard to get funding. And Wellesley College has obviously a great economics presence. Economics is the largest major on campus, and there are wonderful economics students to work with, wonderful faculty to collaborate with. And there’s a lot of demand for this kind of work right now.

This article, contributed by Susan Lowry Rardin, was made possible through support from the Mary Joe Gaw Frug Fund.