SPEAKER: The following is a presentation of the ILR School at Cornell University. ILR-- advancing the world of work.
LOUIS HYMAN: So good evening. Welcome to the second season of the future of work, here, at the Cornell University's ILR School. I am Louis Hyman I'm a professor here at Cornell. I am also the director of the Institute for Workplace Studies, which is the Center for the Study of the Future of Work at Cornell University. And like you, I have many questions about the future of work that I think should be answered with something called data, facts, reality.
What's up with that? Like how big is this gig economy. Who works in it? Why do they work in it? Why is it so hard to know what's going on with all those Uber drivers and delivery people GrubHub, or even how many people simply work a freelance job? And you think to yourself, well, doesn't the government track that? What's wrong with those people? Why don't they just tell us what they're doing? And so tonight, we're going to get to the bottom of it. We're going to find out exactly what's going on. And we have someone with us tonight who, for labor economists and for data nerds is a celebrity.
Well, for regular people, you think, oh, Ariana Grande, you think, The Rock. Well, for us it's Erica Groshen. And we have her here tonight. She was the Commissioner of the Bureau of Labor Statistics until just a few months ago, which is that government agency that is responsible for all of those number that tell us exactly what the workforce is doing. Or at least, that's what they try to do as much as humanly possible.
Now, in her own research as an economist, she's worked on all the important issues like jobless recoveries, male/female wage differences, wage volatility, internal career paths, and other similarly relevant questions about our contemporary economy. So while we call here a data nerd, she's actually someone asking the deepest, most important, most crucial questions about how we work today.
And while she was at the Bureau of Labor Statistics, she pushed for better numbers around this workforce. And she'll explain tonight what we know, what we don't know, why we don't know what we don't know, and what we can do about it in the next couple of years. So please welcome Erica Groshen.
ERICA GROSHEN: Well, thank you, Louis, for that warm introduction. I'm, as many of you might know, I'm actually visiting the ILR School this year. So it's both, an opportunity to get to know better all of the work that's taking place, here in the ILR School, and I really think it's going to be an interesting and productive year. Because you know, the labor market is the largest, the most complex market in our economy. 60% of costs in this country go to labor. So it is the biggest market.
And you know it's the most complex because an hour of labor is one of the hardest things to define-- what it does and what it's worth. And the ILR School, where you are right now, is really the broadest and the deepest assemblage of people who try to understand that, probably, in the entire country. This really is a premier place for understanding that. And right here, in New York City, this is the arm of the ILR School that reaches in to talk to the people of New York City. So it's a special place to be. And I'm honored to be able to be here this year.
What I want to do today is talk to you about this question that people have been asking, certainly, the whole four years I was at the BLS people were asking, what about the gig economy? You know, what are you doing to count the gig economy? How are you doing it? Why do we know? And what I want to go through is, first of all, you can't count something without knowing what you're counting. So what are the gig jobs? What does that term mean?
Let's define it so that, at least in our conversation today, we'll know what we're talking about. We'll then, talk about why it is we'd want to measure them, what we know so far, what we'll know soon. So although, there's some things we don't know yet, we're going to get some crucial information very soon, and talk about what else we need to know. And I'm actually then going to tell you how you can help. Because there's an important role that you can play. So let's get going.
I want to spend a little time talking about the BLS mission. BLS is the Bureau of Labor Statistics. And the reason that I'm doing that is partly because, well, old habits die hard. I'm used to starting my talks by telling people about the BLS. But even more than that, so that you can understand how important it is to measure things like the gig economy to satisfy the mission of the BLS.
So the BLS is this agency that was started actually-- I often ask this-- anybody know-- nobody from the BLS is allowed to answer this question. I have some my former colleagues here. But anybody know when the BLS was created? I'll give you one clue. It's the--
ERICA GROSHEN: OK. It was the first statistical agency created in the US. And it was earlier in 1920 or 1925 or that.
LOUIS HYMAN: Don't worry. You're not being graded.
ERICA GROSHEN: All right. So it's 1884. And the reason is because this was a time of huge industrial unrest. Employers and nascent unions were killing each other in the streets. And the policy-makers of the time decided that they'd be one step closer to restoring peace in the country if the two sides were working from a common set of gold-standard facts. So the BLS was created to measure labor market activity, working conditions, price changes, the kinds of things that employers-employees were disagreeing about.
It collects it, analyzes it, disseminates this essential economic information. It started off in the Interior Department because there was no Labor Department. Then, when the Commerce and Labor was created, it went there. And then, when labor and commerce were split up, it went to the Labor Department. So in some ways, it is the first part of the Department of Labor. In fact, much of the Department of Labor was created as spin-offs from the BLS. Over time, it was appreciated that the statistical mission had to be very separate from the policy mission, and that was well understood early on.
But why is it so important? If you're a data nerd, sometimes you forget why it is you like data, why you like it so much. It's a very simple answer. Because it helps you make good decisions. That's why you like it. Because you feel more in control, and you feel like you're doing the right thing for yourself when you have data you can rely on.
And so the mission is to support public and private decision-making. We, in this country, want to make good decisions. And we think that people ought to be able to make good decisions for themselves. But the truth is, they can only make good decisions for themselves if they have good information to base those decisions on. So whether it's families or companies or policy-makers, they all need this information.
So what makes information good? Well, think about data. You've got to make a decision on it. First of all, it's got to be Accurate. You've got to get it right. It's got to be Objective, free from bias. You can't worry that somebody is fussing with it. It's got to be Relevant-- the information you need. It's got to be Timely-- get to you in time to make those decisions.
And it's got to be Accessible in the literal sense of the word. You can find it, get your hands on it. And the figurative sense of the word-- you can understand what it really means. And if you put those together, they spell out AORTA, the largest artery in the human body. Data is the lifeblood of the economy and, really, a public good.
So think about it-- one of the things the BLS does is create state unemployment rates. Why does it create state unemployment rates? Well, a lot of money is doled out to states on the basis of how their economy is doing or not. And if each state was in charge of estimating its own unemployment rate, you might get some funny results, right? And they might not really be comparable.
Social security benefits-- if Congress had to fight every year about what to do with social security benefits, it's not clear they'd even get what little done that they're getting done now, right? But this is an important BLS charge. If the BLS makes a 1/10 of a percentage point mistake in the CPI, the federal government will overpay or underpay social security recipients by $1 billion. so that's the charge of the BLS.
Now, let's turn, then, to the question at hand tonight-- gig jobs. You came here because you have a sense of what a gig job is. But you probably also said to yourself, yeah, but what's in and what's out of that bucket, all right? What does it really mean? This is a really vague term. Everybody uses it because it's evocative, and because it's short, but it has many different definitions. There are many different components that people might be thinking about-- short jobs with many clients, short assignments from a platform, short jobs done on a platform, flexible scheduling, all of these things.
I'm going to use it in a particular way tonight to mean the sum total of all of that, just nonstandard work, work that is not wage and salary done for us for a single employer that's going to last for a long time. So when you think of what's a real job, a standard job, this is everything that's not that. And I'm going to do that, but I'm going to divide it up in a special way that was devised at the Bureau of Labor Statistics in the mid-1990s to try and get a handle on this nonstandard work.
And there are two basic components that define what it is that makes work nonstandard. One of them is that you have an alternative work arrangement. And that's jargon that means that you don't have this single boss. You may have many bosses. You may have no boss. So the alternative work arrangement is you've got something other than this clear boss who's paying you a standard set of benefits in a way that is pretty standard across the economy. So that's one part of this.
The other part that people have in mind is that some people are contingent workers. And the essence of a contingent worker is that your attachment to that job is temporary. You do not expect that you're going to be doing this for the foreseeable future. So a temp help job would be in that. All right. So when you talk about alternative work arrangements, you're thinking about self-employment, independent contractors being contracted out to a different company.
When you're talking about a contingent worker, it's that this may disappear. Now these are not mutually exclusive. You could be one, the other, both. When I'm talking about gig work tonight, I'm going to say you're falling into at least one of these categories. When I get to specifics and start to count people, then I'm going to define it much more carefully what I'm talking about. But the gig work I'm using is sort of this catchall category.
So why is it that we have this sense that it's growing? Well, there are a number of reasons why people think that this whole area is growing. One, is that we think that IT has really facilitated growth of this kind of work through the matching ability of IT. Another is we think that the fast pace of change in our economy requires employers to be more flexible than they used to be, more trade, more competition, more technological change.
We also think that this is a way that employers have found to get around employment regulations and some of the costs associated with having regular employees. There's the story out there that new generations may prefer gig work. And there's also just the growth that there are always some industries where there was gig work. Those have tended to be growing faster than the industries that had less gig work. And if I've left something off of this list, I hope you'll tell me because I need to keep growing the list. But there are all these reasons why people have this sense that it's growing.
With the sense that it's growing, come the concerns, all right? And the main reason that people are concerned about this growth in gig work is because it suggests that we're entering a world where risks are being transferred from employers to workers. So we now have diffused or no employer responsibility at all to protect workers in ways that regulation required.
So in a standard job, employers have a responsibility to protect workers by providing a safe environment, by paying them in a way that's regulated, by not discriminating against them, not harassing them-- things like that. When there's no employer or multiple employers, there's no one taking that role, so the worker bears the risk of these things.
There's also, there are concerns about there being less stability for workers, for their families, for their communities, and even the financial system when you get more variable jobs and hours, when you have less social insurance, like unemployment insurance, workers compensation, social security.
And when you have few or no employer-provided benefits-- health benefits, retirement benefits, things like that. So it's a transfer of risk. And also the training role-- when there's a long-term relationship between the employer and the employee, then there's more of a responsibility borne by the employer for retraining workers and training them as they move up in their career.
So for all of these reasons, people are very interested in this change in the labor market. There are some very interesting books. The Fissured Workplace is one of them, but many, many other books that talk about what it is, these challenges. And then there's an interesting paper by two ILR graduates, Seth Harris and Alan Krueger, saying that maybe we need another classification for-- there are employees, but maybe there's another classification. I think they call it independent workers?
LOUIS HYMAN: Yeah.
ERICA GROSHEN: Independent workers, yeah. So what kind of protections can be offered in that case? All right. So that's behind, so why do we want to count gig jobs. This handsome fellow is Baron Kelvin of thermometer fame-- Kelvin degrees, you know?
And around the time that the BLS was formed, he, in much more flowery words, said something like, if you can't measure it, you can't improve it. He was arguing that we needed to measure the things that we have to work with. And that's still true today. So why do we want to count gig jobs? Because we want to assess the size of the issue, and because we want to inform policy and the public about what's going on.
So what do we want to measure? How much there is, what the trends are, what are the characteristics? So who's being affected-- the employers, the employees-- what kind of impacts, and, particularly, the terms of employment so that we really understand them? What's the compensation-- high-wage jobs, low-wage jobs, hours, benefits, training, and the implicit and explicit expectations, particularly, for the duration of work.
How long is this attachment going to be? So that's, if you're going to measure this, what you want to be able to get your handle on. And I'm happy to say that the BLS started measuring these sorts of things beginning in 1995. It was a survey conducted in a number of years-- '95, '97, '99, 2001, 2005-- over 100,000 workers each time, carefully tested questions about their working arrangements and their satisfaction with them. And I even show you there, the cover page of the last one.
But you might notice that 2005 is kind of a long time ago, right? That was the last time that BLS was funded to do this survey. And while I was at BLS, every year, we asked Congress to fund BLS to conduct the survey. And Congress decided not to fund BLS to do the survey. But there's really a crying need for this information. So I'm going to talk to you about the next one. But first, let me tell you about what we knew in 2005.
So the survey focuses on people's primary jobs. So if you're looking just at primary jobs-- now I'm going to divide up these jobs into those two kinds of categories that I was talking about-- the alternative working arrangements categories. And say, OK, what percent of all employed workers fall into these categories? What percent of all workers, in 2005, were provided by contract firms.
And you can see that it's about a 1/2%. Work at temporary help agencies, about 1%, on-call workers closer to 2%, independent contractors were the largest part, 7 and 1/2% of employment. And if you add those up, which you can, then you get to about 11% of the workforce was, by these alternative work arrangement measures, part of the gig economy.
The other measurement, of course, is contingent jobs. Now, this is shown in a different way. Because how temporary your job is is sort of a continuous variable, right? It's not either/or. So depending on how stringent you want to be, then you can count more and more people as being temporary or not temporary. So in terms of all people who are in jobs that aren't expected to last-- over 4%. But if you want to narrow it down to people who had the job for less than a year and expected it to last less than a year going forward, then you're getting down to more like 2%.
These numbers aren't huge, but they're not tiny either. That gives you an idea of where we were in 2005. OK. Also, we have information on who these folks were. So you look at the no-boss folks, right, the alternative work arrangement folks, what you find is that, once you break those categories up, the demographics and the satisfaction vary a lot by what kind of arrangement you're working under.
It's not the same to be going an independent contractor as to be a temp help worker. So look at independent contractors, here, these workers are more likely to be-- than people in traditional jobs-- are more likely to be older and white. So many of them tend to be skilled and high-paid.
Temporary help agency workers are much more likely to be young, female, black, or Hispanic. Independent contractors like what they're doing. 82% of them preferred to a traditional job. Temp help workers-- only one third of them. So very different, depending on what kind of arrangement you're talking about.
Contingent workers-- those who don't expect their job to last-- are much more likely to be young than non-contingent workers. And they're less likely to be white. If they're young, they're more likely to be enrolled in school than non-contingent workers. So a lot of these contingent jobs are taken by students. And about 45% actually kind of like it. They prefer it to a permanent job, at least, for their needs at the time. So you get this sense that it's a mixed bag. But there's a lot of good information here, understanding what roles this is filling in society.
Now, 2005, as I've said, is kind of a long time ago. Does that mean we have no information since then? Not exactly. Every month, the Bureau of Labor Statistics does the current population survey. This is a survey that is the source of the unemployment rate for the country. It's a survey of 60,000 households every month, asking questions about the labor market activity.
And there are some questions in this ongoing survey that are not as detailed and focused on the gig economy as the ones I just showed you, but have some relationship to gig work. So people are asked, are you normally a part-time worker? Are you self-employed? Do you hold multiple jobs? All of these things, you'd think, if there's a big burgeoning of the gig economy, you could see pick-up in those categories, right? So that gives you an idea of what might be going on. So let me show you something very interesting.
Usual part-time, multiple job-holder and self-employed from 2002 to 2017, as a share of employment in the CPS. Many more people, part-time workers-- so let's start with the blue line. That's the top-- starts around 17 and 1/2%, turns down slowly, then the recession comes, big jump up. And since then, been trending back down. Certainly, no huge jump. Next is self-employed, the green line. Around 6%, 7% trending-- not up-- down. And multiple-job-holder essentially flat during this time.
Most people find these lines really surprising. None of them is really aimed exactly at gig work, although, self-employed we do consider in that role, but it's only a part of it. But this is surprising. So what's going on? We have this perception. These numbers are supporting it. Whoops. Let me go back. That's right. And that, I was going to show you something from a different data source, which is the Current Employment Statistics Survey.
And this is a survey of employers and asked how many people do you have employed during the pay period that contains the 12th of the month. So this is the source of the jobs counts that we get on the first Friday. If you look just at temporary help agencies' employment, you see the big dip during the recession. You see the recovery afterwards. It looks kind of flat after that, maybe a minor trend up. But again, no major explosion of jobs.
So how does it square with our perceptions? There are some explanations out there for this, and they're worth going through. I think they're probably all partly true. Let me start. First of all, remember we're working off a small base for a number of these categories. The changes-- we could have doubling or tripling. But if you start from a small base, then the noisiness of the data could easily be masking what we're seeing. And it's really not that many people yet. So some of these changes just may be too small to pick up.
Another part of the explanation is certainly that we have more of this internet matching, but they're the same jobs. So think of so many of the Uber and Lyft drivers. They were taxi drivers before. So it was, for many of them, it was gig work before. Now, it's for a different employer, it has an electronic matching component, but you're not going to see a big change in these measures of the job. Because essentially, it's not some different. That's probably part of the story.
Another part is that we're looking in the wrong place. I haven't talked about some of the other categories of work. And maybe the growth is in those, like independent contractors, and also, particularly, workers that are contracted out to other companies-- so the security guard that works for one company and is placed in a hotel. And then, there's misreporting and mismeasurement, that our ways of gathering this information need to keep up with the times.
So let me start with the misreporting and mismeasurement. There are some recent studies that suggest that gig work is underestimated in the CPS. Larry Katz and Alan Krueger have a study where they took the questions from the contingent worker survey and from CPS, and they repeated it in another kind of sample.
And then, they probed more to see whether these questions would pick up gig work. And looking just at contingent work, they said, well, probably missing some of the action there. But the real number-- our better number-- doesn't look that much higher than what the Workers Survey would pick up. And most of the increase there is actually offline, not. Online and because it was a small base, it's really a small number of people. So not so big on the temporary job part.
But the alternative working arrangements, they found a much bigger discrepancy in what the surveys picked up and what they were able to glean from their newer questions. And on the basis of their back-of-the-envelope calculations, they think maybe over 90% of the post-recession job growth gains were in these alternative work arrangement jobs. I think that the CPS is missing a lot of the growth in self-employment, in particular.
Katharine Abraham, former commissioner of the BLS did a lot of research in this area, has also been working with Susan Houseman of the Upjohn Institute, and some people at the Census Bureau-- a very interesting set of studies. And they've been finding out that many self-employed people report themselves as employees, and that many fail to report non-employee work at all, that the questions that are asked don't get the answers that we expect and that half of temp help workers identify themselves as regular employees in the CPS. So this suggests that mismeasurement is part of what's going on.
Another part is looking in the wrong place. So if you go back to the line I showed you about what was happening to temp help from the current employment statistics survey, the Payroll Survey, that's just one segment of a larger industry classification called professional and business services. And an awful lot of those industries contract work out to other employers. And you can see that this, if you look at this larger area, here, you see some fairly rapid growth as a proportion of the economy-- from 12 and 1/4% up to 14 and 1/4% during this period of 2002 to 2017. So this kind of contracting for work between companies is clearly a growing part of what's going.
So I've been hinting that maybe we're going to be able to get to some of the bottom of this. The Contingent Workers Survey was refielded in May 2017. One-time funding was found by the Department of Labor. Previous administration found one-time funding for it to be able to do this. This Survey repeated most of the questions from the past so that there is continuity, that you can compare what you had then to what we have now.
But two new questions were added. I'm going to show them to you in a minute. The results are going to be available in early 2018. I don't know exactly when, but this is the link to it. And as soon as they have a date for when they're going to release it, they'll, at least, put that up on the website.
Let me tell you a little bit-- so it's going to be very interesting to be able to compare these things and to look at the two new questions. These two new questions, which are going to be asked, both about your main job, and your secondary job are-- is that an in-person job that was obtained through an internet-based or mobile platform? OK. So to try and get at the role of computers in this-- And then, the second question is, is your main or your secondary job a series of online tasks completed through internet-based companies. So the first one, an in-person job is like Uber, that is, mediated through an electronic matching platform.
The second one is like Mechanical Turk. Any of you can sign up for Mechanical Turk, and you can get small amounts of money for complete completing fairly tedious tasks. So the idea is to try and measure how much of that activity there is. Yes So this will be very interesting. Now, when we get this survey, we're going to know a lot more about the employee side, but we still aren't going to know very much about the employer side.
And to really understand this change in the labor market, we need to know about the employer side. We need to know how their characteristics relate to their labor supply choices-- whether it's their size, their industry, their location-- how they assess their options. And the reason that we need to know these things is because, if we want to think about how policy should react to this change in the labor market, we need to understand how it's operating so that we can improve policy effectiveness, so that we can understand what the pressure points would be and any policy change, and how to reduce unintended consequences so that we do only what we want to do and not what we don't want to do.
And also, having information like this will just inform employers and employees about what's going on so that they can make good decisions for themselves at the same time. And the way to do this, of course, is to provide gold-standard data, data that people can trust, that lives up to the standards that we talked about in the very beginning. There is, at this point, no plan for something like this, but there should be. OK.
And even on the worker side, we need to know things that aren't going to come out of this version of the Contingent Worker Survey, right? We need to know more about people's non-main jobs than the contingent workers survey is going to be able to obtain because it's really focused mostly on people's main jobs. We need to be able to understand better how to elicit full answers about activity. Because I've already told you that we have reason to think that there are going to be some limitations on the quality of the data.
And we need to understand what role these jobs are playing in people's work lifecycle, their whole work life. This Contingent Worker Survey is a cross-sectional survey. It doesn't ask about people's past or future. It's just what are you doing right now. And why do we need to do this? Because this is very similar to the kinds of things I was talking about before. We need to improve our incidence measures to really understand how much of this there is, and we need to understand its impact on workers.
And doing this is very straightforward. We need to do the Contingent Worker Survey regularly. And we need to have the bandwidth to update it to change the questions in the way they should be changed and to be able to merge this data series with other data series so that we can get at some of the longitudinal questions and so we can compare this, for instance, with 1099 forms and things like that, so that we can validate our findings and get more detail without adding to the burden of the respondents.
So now, here comes your part.
Statistical agencies don't operate in a vacuum, right? I hope everybody here is thinking about, oh, what is it that I really need to know about what's changing in the labor market? And are they asking the question, the burning question, that's really important for my purposes on the job or in my life?
When you share those needs and those questions and those insights with the statistical agencies, like the BLS, than they have a better chance of using limited resources to do the right thing. So it's this back and forth, this conversation with the user community, that helps to ensure that the products do answer the questions properly.
So get to know your local BLS people. There are five of them here today.
LOUIS HYMAN: Raise your hands.
ERICA GROSHEN: Raise your hand. OK. Marty Kohli, raise your hand. Marty is the head of the New York BLS office, right? Get to know Marty. He's your new best friend. And he can help you get the answers to any questions you have, get you the data that you need. And you can tell him what it is you need to know. OK. The next thing you need to do is make sure that the data that you work with is as good as possible by promoting participation in BLS surveys. They are voluntary. Every household that's in the survey, every company that's in a survey makes their own decision about whether to participate.
BLS gets the highest response rates in the business thanks to the work of these folks here, but it's not 100%. It's often in the 80s and the 90%, but it would be even better if it were closer to 100%. Because every time there isn't a response, BLS imputes the answer. BLS is very good at imputing answers, but the truth is better. You can do something about that. You work for companies. You study in schools. You have friends and family who get invited to participate in these surveys. You're a trusted voice. They need to hear from you that this is really important.
Find out if the organization that you work for participates in BLS surveys. And if they do, thank them, tell them how important it is, and if they don't, get to work on them. Marty can help you with that. It's really important. A lot of financial institutions and educational institutions do not participate. Get after them. A very major educational institution is now participating again in surveys because I reached out directly to them and shamed them into doing it, right? But you can do that too. You can do that from inside it, right? You're a trusted voice-- do it. OK.
LOUIS HYMAN: Was it us?
ERICA GROSHEN: I would be violating a law if I told you.
LOUIS HYMAN: Oh, wow. Well, I would never want you to violate a law, Erica.
ERICA GROSHEN: That's right.
LOUIS HYMAN: That's right.
ERICA GROSHEN: OK.
LOUIS HYMAN: Particularly on camera.
ERICA GROSHEN: You should also champion official data whenever you have the chance. If you use the data, then any time someone impugns that data, they're impugning your work and your trust in it. It's very easy to deflate the cheap shots, the sloppy work, the nihilism that causes people to just dismiss statistics. They don't know what they're talking about. They don't understand the techniques used. They don't understand why it is the data should be trusted.
You need to explain that to them, that this is work done by career civil servants, it is not subject to political manipulation. It is state-of-the-art work. It is the best information out there. You can do more good for the statistical system by the conversations you have at the cocktail parties and barbecues than many of us can do from our lofty perches because we are not the trusted voices that you are. It's fun. I guarantee you will enjoy it because those loudmouths don't know what they're talking about.
But if you let them keep on saying it, they're going to keep on saying it because it gets the laugh, or it makes them seem smart or something like that. And if you don't know how to defend the data, go to the website. There's plenty of ammunition. Or talk to your new best friend, Marty, and he will help you. OK. And then, each of you who works for an organization that has a leg affairs staff, and those leg affairs staff are communicating their needs to the government.
You need to make sure that they understand that you need the data to do the work right. Because they're used to going to Capitol Hill and talking about taxes and talking about regulation. They're not used to going there and saying, by the way, don't cut the budget for BLS. Because if you do, our company won't be able to locate its stores well anymore, or make good financial decisions anymore, or give good career guidance anymore.
They need to hear that from you. It's actually, probably the easiest thing you could ever ask them to do. But nobody's been telling them to do that. And your CEOs need to hear it from you too so that it happens. You need to endorse the budget for the BLS and data-sharing legislation so that the agencies can share information and make the statistics even better.
And just today at 11:00 AM, the Paul Ryan-Patty Murray commission issued a set of recommendations on how to improve official statistics and evidence-based policy-making. Every one of those recommendations needs to be enacted. And it will be the work of the organizations that you work for in supporting this that will make a difference. So I hope you'll find out about the Ryan-Murray commission, and maybe you'll have a session on that one next.
LOUIS HYMAN: Maybe we will.
ERICA GROSHEN: That's right.
LOUIS HYMAN: I'd love to have those people here. That would be lovely.
ERICA GROSHEN: So you can do all of these things. I recommend-- I actually tell you, a lot of them are fun to do. I spent four years doing some of them, and they're fun. OK. So bottom line-- when you think about the gig economy, make this kind of division in your mind between the contingent workers who are in temporary situations. I would maintain some skepticism about a vast explosion of those kinds of workers. That's the one that's hardest to pick up in the data.
On the other hand, the alternative work arrangements where there's no clear boss, that seems like it's growing faster, and especially among independent contractors and contracted out workers. The 2017 Contingent Workers Survey is going to advance our understanding of a lot of recent trends. Keep your eyes open for that. But other information is really still going to be needed on the employer side, more information on the employees side, and the ability to match this information with other information big data sources, IRS information, so that we can really push ahead in understanding this.
And with that, I want to thank you for your attention, and I look forward to all of your questions.
LOUIS HYMAN: So thank you so much, Erica. That was a wonderful talk. And I'm sure we have lots of questions. I have a question.
ERICA GROSHEN: No.
LOUIS HYMAN: And since I'm on the stage, I get to ask first. Then, I'm going to ask my question.
ERICA GROSHEN: Go right ahead.
LOUIS HYMAN: Go right ahead. So what's interesting to me is that what makes this so hard to measure doesn't seem to be the numbers. It's not math that makes it hard, it's words. And it's words, both, in how economists and social scientists are constructing these categories that's difficult. What is this? It's so easy to say do you have a job? And if you have a normal quote unquote, "full-time permanent job," they're, like, yes, I have a job. But otherwise it's much more amorphous.
So there's that challenge of words. And then, there's a challenge of words of people who are actually working in these ways. So I did this survey last year. And part of the survey had all these Uber drivers in it. And you look at-- they say, I drive for Uber, I'm unemployed. I drive for Uber, I have a part-time job. I drive for Uber, I'm a full time independent worker. And I'm a full-time employer, who's my boss? Uber. And part of it is the words we use to describe ourselves and our identities.
And this seems to be one of the great challenges. How do you think this can be resolved, going forward with these kinds of survey methods, which are, themselves, based on a more stable notion of what these things are, which, right now, things are very unstable? How do you grapple with that as a social scientist?
ERICA GROSHEN: It's a really important question. But it speaks to the expertise that you find in the social science community and, in particular, in the statistical agencies. They are experts in trying to figure out exactly how to ask these questions so you understand the answer you're getting.
There is an entire process that they go through to design the questions to make sure that they're only measuring what can be measured and that everybody understands what it is that they're after, whether it's-- and you need everybody to understand. You need the respondent to understand, you need the interviewer to understand, and then you need the analyst who gets the data to understand. And so BLS has a cognitive lab where they--
LOUIS HYMAN: --do it.
ERICA GROSHEN: Yeah. Oh, yeah. And they go to all of these steps that you need to devise questions. So they start out with focus groups, and then they do testing, and then they do retesting, and then they'll run parallel surveys.
LOUIS HYMAN: So this is what you mean by gold standard. It's not like when I run a Google survey that I just mail to my friends from high school. I'm like, hey--
ERICA GROSHEN: That's right.
LOUIS HYMAN: Are you-- drive for Uber?
ERICA GROSHEN: That's right.
LOUIS HYMAN: Right? Exactly. OK.
ERICA GROSHEN: So that's one part of the answer that, for it to be a gold standard, you need that background work. Otherwise, you don't know what you've got. And even when you have something, you can't be sure that it continues to be relevant if you don't keep testing it. So that's part of it. The other part of it is that along the way, you have to define what these buckets are.
And so that's some of the work that the BLS did, and say, when they designed the Contingent Worker Survey, they did the background work to say, look, these are the two key buckets-- the alternative work arrangement bucket and the temporary bucket. And we have to talk about them separately. They are different concepts, even if we often conflate them in our normal conversation because they really are separate.
LOUIS HYMAN: Because there's just so much we've talked about, we'd say, just not normal.
ERICA GROSHEN: Right.
LOUIS HYMAN: All the other stuff-- not normal-- whatever that normal is. We have a lot of questions from the audience, and some online. So I'm going to take some questions from the audience. And then, we can go to online. First-- yes, thank you.
LOUIS HYMAN: Yes. And then-- just wait a second for the microphone.
AUDIENCE: Thank you. Hi, Erica. Great job.
ERICA GROSHEN: Thank you.
AUDIENCE: So My name is Marjorie McFarland. I learned a lot more about the gig industry, or the gig economy, over the summer in a program I took at Cornell. And I actually decided to start a gig company. I landed my first contract-- woo-hoo-- a month ago. And we're charging forward. So my question for you-- in your slide presentation, you mentioned that your surveys were done starting in '95, and the last one in 2005. Which industry did you conduct these surveys in, and why? Thank you.
ERICA GROSHEN: This was a nationally representative sample. So every industry in the economy is represented in the survey. OK. And that's intentional to say where is this happening?
LOUIS HYMAN: Another question? Gentlemen back there?
AUDIENCE: Hello, Erica. Great job. My name is Jose Torres. I work at the BLS. I'm an economist. I work at the Employment Cost Index Occupation Requirements Survey as well as the Employee Benefits Survey. And I love when you mentioned how 10 basis points in the CPI can cost the US government $1 billion, yet, to BLS budget is only $609 million a year. Personally, I see that to be such a high risk, to only fund the BLS budget by $609 million. How do you feel about that?
And what do you think we can do as American citizens, or as BLS employees, to let Congress know, listen, we're taking a big risk by underfunding this agency and only funding it by $609 million. Yet, a small issue with the CPI can lead to such a large expense. And that's only the CPI, not to mention the ECI, or the unemployment numbers, or anything else. How can we emphasize that to our leaders in Congress? Thank you.
LOUIS HYMAN: And the CPI is the Consumer Price Index, which is the measure of inflation.
ERICA GROSHEN: So you ask a great question. BLS employees, of course, as citizens, can vote like any other and have conversations with their friends and families. But officially, they have to step back from that. However, everybody else here who is not a BLS employee needs to let their elected representatives-- because they are the ones who are in charge of transmitting your will to Washington, to make the decisions on how the budget is allocated-- they need to know that these data are really valuable.
The US has not had an inflationary increase-- and it's had an increase in prices that it pays and in wages to its workers, which is a good thing. The prices-- wages have continued to go up, and BLS's budget has been flat. It's, in the past seven years, that's been a real cut of 14% in BLS's budget. If it's flat again this year, BLS will have to start cutting programs. It's already at risk-- with certainty, it's falling behind in modernization. And the risks of a failure of the employment situation not coming out on time, like the CPI not coming out on time because outdated hardware or software failed, and there isn't enough bandwidth to cope with the crisis, is just rising every day.
So these leg affairs folks who represent your company's interests need to speak up. And you, as citizens, need to get in touch with your congressmen and senators because they are making those decisions right now. Today, the Senate Health Committee voted on the market for the BLS budget for 2018. I actually don't know what they decided yet. The House voted flat funding again. Senate, I don't know. And too many people are saying oh, you're lucky it wasn't a cut.
LOUIS HYMAN: Our businesses and our markets rely on this information to make decisions every day.
ERICA GROSHEN: They do. And they have not spoken up for it. They do not speak up for it. And I don't know why. And if you can come up-- that's another thing you can do. If you can come up, if you can tell me how, if we can get them to speak up--
LOUIS HYMAN: Well, there's one more question from online, and then we're going to have to break because we're running out of time. Someone online asked, the GAO produces estimates of broadly defined nonstandard [INAUDIBLE] work. In 2015, using a range of sources estimating about a third of the workforce.
Of course, there's other kinds of surveys-- the Upwork Freelancers Union Survey. People talk about this. What would make this different than those kinds-- why is this better, not to put it in terms that would be causing you to have a brawl with the GAO. I know there's a lot of [INAUDIBLE] and beef between the BLS and the GAO. Why is this better than the GAO or other kinds of surveys?
ERICA GROSHEN: Well, the biggest difference between the GAO estimate and the estimates that I've shown you is that, in their view of nonstandard work, they decided to include all part-time workers as nonstandard. A definition they chose, they're very transparent about it. If you add in all part-time workers, that number gets a lot larger.
So that's the main thing. They chose a very inclusive, very broad definition. But a lot of part-time workers have all of the protections that we've been talking about. So the BLS decision was not to include them. But one always could add them back in. There's no reason why you couldn't.
LOUIS HYMAN: So it gets back to these questions of words and where we draw lines. And we're still debating that.
ERICA GROSHEN: There's another subtlety about some of the other estimates of the size of the-- if someone says what's the size of the gig workforce, that's different than what's the size of gig employment. So this is a subtlety about words. But when people estimate the size of the gig workforce, they want to say, over the course of the year, what's the share of people who never held a gig job? That's a much larger number than in any particular week how many people were working in gig jobs.
So one is saying, what share of the productive activity is kind of flowing through gig jobs. So that's more that's more the BLS concept. This size of the gig workforce is one-- it's a legitimate question. But you would ask that if what you're interested in is how many people have an interest in that work. Or if you were, say, a gig employer, and you wanted to know how many people you might be able to attract to do work for your company for small amounts of time or large amounts of time, then the size of the gig workforce would be more what you want to know.
LOUIS HYMAN: Well, thank you so much. This has been fascinating. Before we thank Erica Groshen for coming tonight, to hopefully welcome you to the next event, next month. We're going to have Michelle Miller who is the co-founder of coworker.org, which is the first and largest digital worker organizing platform for bringing together people who work at Starbucks and all those other kinds of part-time and shift workers that may or may not be-- depending on who you ask-- part of this contingent workforce.
But thank you, Erica Groshen, for coming tonight and explaining to us the importance of data. I hope we all do reach out to our representatives to get the truth out there in the world. And thank you, everybody, for coming.
ERICA GROSHEN: Thank you, Louis.
LOUIS HYMAN: And we're going to have some more--
SPEAKER: This has been a production of the ILR School at Cornell University.
We've received your request
You will be notified by email when the transcript and captions are available. The process may take up to 5 business days. Please contact firstname.lastname@example.org if you have any questions about this request.
In the latest installment of the Future of Work series from NYC, ILR Associate Professor Louis Hyman is joined by Erica Groshen, former head of the United States Bureau of Labor Statistics (BLS), 2013-17. The last time the BLS surveyed the contingent workforce was in 2005—this year they will do so again. A lot has changed since 2005, and we expect there to be a very different result this year. Dr. Groshen explains why it has been so hard to count workers in this new economy, and what that means for how we think about economic policy.