Bill Kerr: The U.S. Bureau of Labor Statistics is considered the gold standard for economic data. The agency tracks labor market activity, working conditions, productivity, how people spend their time, and much more. Its closely watched reports inform public and private decision making at every level. They also have the power to ignite political battles. Politics has always threatened the BLS’s impartiality, and in 2021, the agency is emerging from an especially challenging four years. At the same time, it faces the need to adapt to recent changes in the world of work. How can this source of critical data and analysis maintain its independence and relevance?

Welcome to the Managing the Future of Work podcast from Harvard Business School. I’m your host, Bill Kerr. My guest today, labor economist Erica Groshen, was commissioner of the BLS from 2013 to 2017. While there, she led efforts to expand the agency’s data gathering and to increase coordination with other federal statistical agencies. She’s currently affiliated with the Cornell School of Industrial and Labor Relations [ILR] and the Upjohn Institute for Employment Research. She advocates for enhanced BLS reporting and a more rationally organized and independent federal statistical research system. We’ll talk about that and the prospects for a pendulum swing toward greater worker representation. We’ll also discuss workforce development, gig work, income inequality, the social safety net, and the impact of the coronavirus pandemic. Welcome to the podcast, Erica.

Erica Groshen: Thank you. It’s a pleasure to be here.

Kerr: Erica, can we begin with a little bit about your career? What was the trajectory that took you to the BLS, and what are some of your current research areas?

Groshen: Well, I started college at the University of Wisconsin–Madison as a physics major. I switched majors to econ and math. And then after graduating, I worked for four years doing programming policy evaluations, and then went to grad school at Harvard. And there, I focused on labor economics, because the labor market, as we all know, is the largest and most complicated market in the economy. After that, I spent 25 years in the Federal Reserve System—first at the Cleveland Fed, then at the New York Fed. And there I did things like publishing research, advising policy makers, participating in regional outreach, managing, and various things like that. And during that part of my career, my research focused on the role of employers and labor-market outcomes, such as wage gaps between men and women, inequality, inflation transmission, and recoveries from recessions. Then, as you mentioned, in early 2013, I became the 14th commissioner of the Bureau of Labor Statistics, BLS. And that’s the federal agency that’s responsible for measuring labor market conditions, inflation, and productivity. Since my term ended, as you mentioned, I’ve been affiliated with the Cornell ILR School and Upjohn. But I’ve also been serving on a number of advisory committees and chairing the Friends of BLS and doing some research and consulting. And lately, my research has been focusing on how to improve our national statistics and also on the impact of the pandemic.

Kerr: Let me go back to the time at the BLS. And one of the things that you pushed for while you were the commissioner was enhancing worker wage records. And so I want you to first off position for us, what are we talking about here? And then second, what are some of the metrics that you would like incorporated into worker wage records?

Groshen: The unemployment insurance system, the national system, is actually a federal system. So you have state unemployment insurance agencies that, as part of running the program, collect worker wage records every quarter from every employer that lists the wages of workers for every month during that quarter. They also collect claims records from people who apply for claims. And these data are generally not available to BLS to augment or replace its current data collections. And that’s basically a shame, because it would be quite useful for statistical purposes. And employers, of course, have to report the same or slightly different data to a number of different government agencies. Our economics statistics are also not as good as they could be as a consequence of this. UI [unemployment insurance] wage records include who the person’s employer is and their earnings—that’s what’s in there. They should have job title, because that is closely associated with the person’s occupation. And through the use of artificial intelligence, you can assign job titles to the standard occupational classification system. And this would enable us to track workforce supply and demand much more closely, make better projections about the future of work. You also would want the number of hours worked for the wages that are being reported so that you know if someone is full time or part time, so you can get hourly rates, and really follow that dimension on which wages vary. Another thing you want is the actual work location of the people, because sometimes the reporting unit for unemployment insurance is not where the people are actually working. And sometimes these accounts cover more than one unit. But if you’re trying to understand the way jobs are distributed, then you would want to know where each person that you’re tracking is located. And then, the last thing, particularly in these times of understanding demographic inequities—racial inequities, in particular, but also gender inequities, things like that—you want to have demographics so that you can track social justice issues and advances and understand how the world of work is affecting demographic outcomes. These data should also, of course, be curated—by which I mean, they have to clean them up so that you can really analyze them and made accessible to the statistical agencies, for particular with the BLS, so that they can create better statistics. You could get better, cheaper, and more-frequent program-policy evaluations so that policy makers could make better decisions. So the results would be better decisions by these policy makers, by voters, by job seekers, by workers, by companies, by investors, by the whole workforce development system. And, as everybody makes better decisions, then the economic vitality of the whole country benefits.

Kerr: So in a time where we’re really looking to enhance the upward mobility of workers, do you see this as a critical input for being able to measure how much upward mobility is there? And then also, what are the things that are enhancing those upward movements?

Groshen: Absolutely. You can look at what kinds of skills are really paying off for workers, and you can model what kinds of opportunities may be coming in the future.

Kerr: If there’s such a great sort of compelling set of use cases for having this type of data, what are some of the challenges that are getting in the way of us improving our data collection to obtain it?

Groshen: One of those, of course, is actually funding. Another barrier is that, in the past, these agencies’ missions really haven’t included creating a national data infrastructure. There has to be some recognition of this role as being important to their mission. I’ll give you an example of one of the things that happens when you don’t have the attention of national statisticians to administrative data. You can think about the unemployment insurance initial claims releases. There’s been a lot of attention to that during the course of the pandemic, because it’s some of the most-timely data and very closely associated with what was going on in the labor market. But those are administrative totals. They are not constructed to be economic indicators. And most of the people paying attention to them were looking for an economic indicator. The solution is clear, which is to have BLS partner with the unemployment insurance system and take over production of the creation of economic indicators from this inputted information a new program that takes advantage of the skills of a national statistical agency to input that data and create an economic indicator that wouldn’t require all of the journalists and all of the economists everywhere else to say, “Well, let me make this adjustment; maybe that’ll tell me what’s really going on.”

Kerr: With the efforts to improve HR data inside private companies, are there interfaces between the public sector and the private sector to really standardize data-collection methods, do it in ways that benefit both the firms and also the government?

Groshen: Yeah, there’s a really interesting effort underway called the “T3 Innovation Network.” This is led by the U.S. Chamber of Commerce Foundation, with support from the Lumina Foundation and a bunch of others. And they’re doing something really interesting, which is designing standards for human resources data that will also be used, ideally, federally for data collections, industry standards for human resources files, for employment and earnings files, and also for skills information that would be interoperable between different companies, so that when they merge or split up, they’d have common data elements, so that when they buy services from a software provider, that these fields would all be in common. And from the point of view of the statistical system, it means that, then, the request for data from a company would simply be, “We want this field, this field, and this field” from the set that already has these standards already in place. And this would help companies reduce the reporting burden, then becomes much easier. It also improves their internal analytics because there will be suites of products that are all built around these common internal analytics. Then the companies will have a much better opportunity for benchmarking, because they’ll know that their internal numbers are comparable to the external numbers that they’re seeing being published. They want to come up with a mechanism to provide workers with portable, authoritative job records that they could tap into for applications, for jobs, for educations, for UI benefits, and other public programs as well. Anytime when they say, “What’s your work history?” the worker would be able to plug in their ID and their password or something like that, and they’d have an authenticated work history with skills information and duration and other information on it. So it seems like a very clever idea whose time has come.

Kerr: Erica, whenever one of the BLS reports comes out, there can often be a fight about the narrative of this between different political parties. How does the large political polarization we’re experiencing as a country right now affecting data collection?

Groshen: Interference—and just the appearance of it—can diminish public trust in the data. And that public trust is actually a critical element for a statistical agency to fulfill its mission, because data providers need to trust that the data they provide will be used safely and properly. If they think it’s going to be used to create garbage or to somehow punish them in some way, then why would they provide it? So on the data-provider side, you need this trust, because [when] the BLS field staff go out and talk to a company, the biggest pitch that they have is, “Look, you’re performing a crucial public service.” Trust is very important on the data-user side, because why produce statistics if people don’t use them? And people aren’t going to use data whose integrity they don’t trust. Political polarization and attacks on data really do undermine the capacity of statistical agencies to do their work. And in a very glaring recent example, the attempts and some of the challenges with the 2020 decennial census taken by the last administration—this is a really key element of our national statistical infrastructure. It underlies so many of our data sets and so many decisions taken on Capitol Hill by companies and everyone else. And these actions expose vulnerabilities in the laws and the practices that are intended to protect the statistical agencies from interference. Polarization can also inhibit adequate funding of things like the statistical agencies, because then their work is seen through a political lens rather than as the nonpartisan good government issue that it really should be. There are many steps we can take to strengthen the independence of statistical agencies. And one of them would be to put them in their own agency outside of the control of a member of the cabinet, have it be headed up by the national statistician of the U.S.—and that’s a job that exists, but right now, the national statistician of the U.S. has an office of about seven people in OMB [the Office of Management and Budget]. So the alternative is put the national statistician actually in charge of the statistical agencies and move them away from reporting to the people who are in charge of policy directly.

Kerr: Erica, we’ve spent the first half of this podcast talking about all the many ways you’ve helped collect data, but your career has also spent a lot of time crunching the data. Let’s start with the skills gaps. Can you tell us a little about factors that are widening the skills gaps among employees?

Groshen: I think a lot of it really comes down to information problems regarding workplace skills. We know how to count years of education, and we have some ideas on quality. And we know test scores and things like that. But we actually don’t track workplace skills very well—either on the micro or the macro level—for individuals or for the country or groups as a whole. A very large component of skills training is done by employers. And we have no good measures of that. We don’t know who gets what kind of training. We don’t know what our skill gaps are. And so how can we be making the right decisions if we don’t have that information? The last authoritative source of employer-provided training was by the BLS, and it was done in 1995. Congress hasn’t funded it since then. And it really should, because if we want to have a 21st-century workforce, we need to have this information. So workforce development agencies, job counselors are making decisions that influence what kind of training people get; job seekers search out information on what kind of training will make them more attractive to employers. And they don’t have as much and as granular information as they could use. One other thing that’s important is the employer side of all of it. And that’s really where I’ve focused a lot of my work in the past. These days, employers are increasingly making their decisions on who to hire and who to promote on the basis of computational algorithms. And this is problematic for people who don’t have these credentials—like college degrees, in particular. So I’ve been participating in some research that suggests that people without college degrees are being excluded from many jobs that they could do with little or no additional training. And this has been compounded by this reliance on computer-assisted applications and screening mechanisms. In particular, this discriminates certainly against displaced workers—workers who have been on the job for a long time, lose their job, and they have to go to a new employer. And there’s no way of really documenting the skills that they obtained on the job. And it also is problematic in a time when we have more and more fissured workplaces, where more and more employers are hiring—not workers, but other companies to provide labor to them. And that means if you want to work up from a lower-skilled job to higher-skilled job, that may mean that you need to change your employer, to move from being the contracted-out worker to being the permanent employee, or to move to a different kind of company entirely. And that feeds into workers fearing innovation and resisting that. And it stifles their career paths. Used to be that in the “olden days,” the secretary was an understudy to the boss. Alexander Hamilton was secretary to George Washington. Those career paths have been separated, and few secretaries get to be CEOs anymore. And that’s happening along other career paths as well. And this is particularly a problem for workers without much formal education, and that Black workers tend to have less-formal education, so it’s a problem for them. And employers are really losing out, because they’re not tapping into a much larger pool of workers than they could be.

Kerr: He’s not here today, but my podcast cohost Joe Fuller’s eyes and ears widen when you talk about poor decisions about hiring based upon college degrees. So I’ll put a plug in for his report “Dismissed by Degrees.” The fissured workplace that you were ending on, let’s use that as a segue to go onto another question, which I’m sure people at the BLS are spending a lot of time on, and also researchers are trying to understand more about, which is: What’s the role of the gig workplace, freelancing? How do we distinguish, perhaps, between a few of those different types of work modes? And then, what do we need to learn more about the effectiveness of them in the economy?

Groshen: The one area where there’s been a real big growth of platform work has been in transportation, particularly Uber and Lyft kind of jobs. So those have definitely gone to this gig work kind of thing. In other segments of the economy, it hasn’t been as widespread. But yet, it is growing, there’s innovation. And the challenge in all is that there’s no standard definition. And this is partly a result of policy used to distinguish between basically normal wage and salary work and everything else. And it didn’t pay much attention to the “everything else.” So we can define who’s a normal wage and salary worker, but everything else doesn’t have a legal definition. BLS has taken an approach with this survey of alternative work arrangements, which is basically everything but wage and salary. The challenges are, because we measure wage and salary pretty well—primarily based on Fair Labor Standards Act coverage and coverage by the unemployment insurance system—for some of the other kind of work, we can look at 1099 forms. But that’s a problem, because 1099 forms don’t have very much information on them. They only come out once a year, they’re lagged. And so if you’re trying to keep track of this on an ongoing basis, it’s not easy. And then, with the fissured workplace, when you’re looking at a particular company or an establishment, if they’re contracting out a lot of their work, you don’t actually know how many people are working at that facility, or how many jobs that activity is supporting, because it doesn’t get reported in any easy way. Easy example of that is temp workers aren’t reported as being employed by the place they’re assigned to. So it’s very hard to measure total labor input for productivity measures. It’s very hard to track labor market activity on a national or even a local basis if you don’t know where people are working or what they’re trying to produce. It’s hard to measure career paths and mobility, because you don’t know how people move within it. And even when you just ask workers about what their activities have been, how many hours they’ve worked in, how much they’ve earned in different ways, you run into difficulties in remembering and even summarizing what they’ve done. And some of these new approaches—like Uber, Lyft, but also Airbnb—mix providing labor and capital. So when you drive for Uber, you’re not just providing your time, but you’re providing your car. And when you’re doing Airbnb, you’re renting out your home the same time as you’re cleaning and servicing the facility. So there are data challenges with all of them. The statistical agencies have dealt with these challenges in the past, and with appropriate funding, they can deal with them again. But BLS has had flat funding for over 10 years now. It’s very difficult to innovate and keep ahead of things when you have flat funding and costs are rising and the world is changing rapidly.

Kerr: And to your earlier points in the conversation, that affects many of our analysis as economists, the quality of inputs we’re giving into policy advice, and similar. Erica, I want to stretch the timeline here and take it all the way back to include your days at the Fed. You had worked, for example, that talked about jobless recoveries and productivity declines. What were the factors you see as being the decline in our rate of productivity growth? Used to be higher, and then for more decades than probably we care to count, it’s been dampened. What do you ascribe it to?

Groshen: People have talked about energy and medical care and those things, and they play a role. They’ve also talked about measurement issues, too. I don’t think it’s about measurement issues primarily. But on the labor side, I think there are some really important questions. For one thing, interestingly enough, the decline in unionization and in worker voice, particularly since the 1970s, I think deserves more attention. We know from past research that unionized workplace are actually more productive than nonunionized workplaces. You can see it in a number of different ways—if only because the wages are higher, and so the employers need to get every bit of work out of those workers so that they can still be profitable. But also workers who are paid more have less turnover, and they’re more loyal. So we know the workers earn more in those settings. And the returns to capital, though, are a little bit lower. So the workers get the benefit of this productivity difference and then some. And so this goes along with the fact that, since the 1970s, workers’ compensation as a share of national output has been declining. So as the productivity growth has slowed, employers—the owners of capital—have gained more and more of the national pie. With more unionization, what you’d expect from this sort of analysis is higher productivity, and maybe some decline … higher productivity and increased share of national output. And you might think that, that would be quite appropriate. It was certainly seen as quite appropriate in previous decades. So it’s not clear why it wouldn’t be appropriate now. One road to improving productivity is more unionization. One way that you can see how this works is that when workers have a voice, we can get back to one of the issues that we talked about before, where workers often have skills that their employers don’t actually fully appreciate. And when you have worker voice, then you have this opportunity for discussion about this, so that the decisions made by the employers are actually with full information. The other element in all of this is when workers—and I mentioned this before—when workers fear that innovation is going to deprive them of a job and a livelihood and a career path going forward, they are understandably more resistant to innovation than they would be otherwise. Where they feel valued, where they feel that they have a voice to help shape the future, then they’re going to be much more supportive of innovation than they would be otherwise.

Kerr: Erica, as you think about this call for greater worker voice, we’ve seen over the last three years this podcast, a range of opinions from, we need to go back to some of the traditional representation models that we had in the United States in the late 20th century over to, we should take this element of the Swedish model or some other European model, bring it over to yet third or fourth variants. What would you be recommending as the best next step for us to look or contemplate or experiment with as we think about this worker voice?

Groshen: I don’t think we know right off the bat, what’s going to be the right solution. There’s more than one way to improve worker voice. If the past is any guide, we’ll probably come up with not only just one way, but a suite of ways that are appropriate under different circumstances. Unions are experimenting, but there could also be experimentation on the corporate side and on the policy side, on incentives to provide it to companies to get that worker voice innovations to institutionalize its studies, and look what this could look like. The studies of German-style works councils have suggested that they’re very important part of management decision making and very useful to the companies in implementing innovations. So that would be an important thing to look at. Japanese companies also have mechanisms that are built in for manufacturing workers—in particular, to be able to provide input in design tasks and work in teams.

Kerr: So, Erica, as we record this, it is late May of 2021, Covid has been with us for about 15 months, and we’re to a point now where most people that want vaccines in the United States have access to vaccines. And yet we’re also looking at various reports that come out. Jobs go up by a great amount, one month, and then they’re disappointing the next month. So I’d like you to just maybe end by giving us a little bit of your perspectives on what is coming out of Covid-19 right now, and what are the big challenges that we should be thinking about for the next five years?

Groshen: One month is not a trend. Whatever month you’re looking at, you should never assume that that’s the trend. You should look back at least three months. But I do think that right now, at the end of May 2021, we are at a time when we have a lot of churning going on. We have a lot of some companies expanding, other ones contracting. And so just looking at the top line is not going to give you a very helpful picture of what’s going on underneath. I recently looked at a wage growth using the employment cost index by different occupational groups, and I found that in the most recent quarter, the variance of wage growth by occupations was as the highest it had ever been measured in the series. We have a lot of churning going on. But I think the big challenges are some of the gaps that have really been exposed. So an easy straightforward one is we have big gaps in affordable, accessible childcare. That’s crucial for women workers to keep people in the labor force. And almost half of our labor force is now made up of women. It’s also important for the next generation of workers, because this is what will give them the best start. And so I think a focus on ensuring that our childcare system—our childcare infrastructure, if you will—is robust is something that will come out of our experience with us. And associated is also our nursing home system, because for some people, ensuring that their parents or their loved ones have good care is as important as having made sure that their children are. I think the work of non-college-educated workers, who are very heavily represented, they’re most of our essential workers is also critical to the functioning of our economy. And that’s been revealed to us in very stark terms. And we need to safeguard their ability to work with things like sick pay and health insurance, and other steps that will mean that they can work safely and productively for us, because the economy will suffer otherwise. Labor force participation of older workers is going to be an increasing area of focus, too. As our population ages—up until the pandemic—older workers was one segment of the population where labor force participation was growing—old people were working more than they had in the past. The pandemic led to big decrease and not as much of a rebound for that segment of the population as for other workers. And so, I think it behooves us to think hard about what’s going on with that set of workers. And reliable high-speed access to the internet turns out to be really important for continuing to function in this world. Jobs, education, provision of any kind of services—our increased dependence on that laid bare the fact that, that is not equitably distributed across our economy. And then finally, I think the experience of the pandemic has shown us really graphically, very starkly, how much we need and could have more granular, timely, and relevant data to guide decisions. So we need an infrastructure capable of producing it. We need the option value of those data, because we don’t know what’s coming next.

Kerr: Erica Groshen is a labor economist and was past commissioner of the Bureau of Labor Statistics. Erica, thanks so much for joining us today.

Groshen: You’re welcome. It’s my pleasure.

Kerr: We hope you enjoy the Managing the Future of Work podcast. If you haven’t already, please subscribe and rate the show wherever you get your podcasts. You can find out more about the Managing the Future of Work Project at our website hbs.edu/managingthefutureofwork. While you’re there, sign up for our newsletter.