Discussion: (4 comments)
Comments are closed.
A public policy blog from AEI
Have poverty levels and inequality in the US soared over the past quarter century, or are we just looking at them through the wrong lens? I discussed this recently with economist Bruce Meyer, a leading researcher in these fields. We also covered the earned income tax credit, how government can held reduce poverty, and more. Bruce Meyer is a visiting scholar here at AEI, a professor at the University of Chicago’s Harris School of Public Policy, and a fellow at the National Bureau of Economic Research.
Let’s start by talking about income inequality. It’s a huge economic issue, but probably more so on the left than on the right. When I think about inequality, I think of charts frequently shared on social media showing the income taken by the top 1%. The top 1% used to take about 10% of the income, and now it’s creeping closer to 20%. Do these charts say that something has gone very badly wrong with the economy?
My work really is not about that part of the distribution. We look at the bulk of the population, those between the 90th and 10th percentile of the income or consumption distribution. So, I think that you should care about the top 1% but that’s a very small share of the population and inequality across the broader population is much more relevant for who you live next to, who is in the car next to you, who’s walking down the street by you.
So why do we focus so much on that number? If someone hears this is a podcast on inequality, they’re going to think, “Oh, we’re going to be talking about the 1% or the 0.1% or the 0.01% and how they are doing.
Well it’s kind of a sensational number, and it is important if you’re thinking about tax policy because a lot of the income and wealth is there. So, even though it’s a small share of the population, it’s certainly relevant for some policy issues.
When people talk about inequality, usually in the next breath they’ll say, “Listen, that top 1% is now getting a lot of the income, and at the same time, incomes for the middle class have been stagnant for 30, 40, maybe 50 years.” Those two facts are used together, I think, to show a relationship between the two. Is that right? Because to me it seems intuitively, how could that possibly be right that the median American is no better off than he was sometime in the 1960s?
I think you’re right to not believe it, I think that statistic is wrong. If you look at the consumption of the median household, it’s gone up a lot over the last 30 years. Part of the explanation for the difference between what I am saying and some of the conventional statistics that you might see from the census bureau is that we often use a way of adjusting for price changes that overstates the extent of inflation. So it makes it look like we are not doing well in the middle of the distribution even when we are.
There are very tangible things that you can look at to see that people in the middle of the distribution are better off than they were 20, 30 years ago. For example, if you look at the share of people in the middle of the income distribution that have central air conditioning, or maybe only a couple of room air conditioners, or have a dishwasher, or a washer and a dryer in their house or apartment — those numbers for the middle look like the numbers for the top 20% as of 20 or 30 years ago. So there’s been quite dramatic improvements if you look at tangible things like what kinds of appliances have in the house. You can also look at the size of people’s houses or apartments — square footage, number of rooms — those things have gone up quite sharply for people in the middle of the distribution.
So if we’re looking at how people are doing on a pre-tax basis, how we measure inflation is very relevant. But do those income numbers include everything?
Well, when you talk about the broader income distribution, particularly those at the bottom, it’s very important to pay attention to taxes, and in-kind transfers, and to the fact that our income data sources tend to undercount a lot of key transfers from the government. Much of what we have done over the past 20 or 30 years to help those at the bottom of the income distribution has been in the form of tax credits, like the Earned Income Tax Credit, or transfers that are in-kind like food stamps, which we now call SNAP, or housing benefits, or Medicaid. And those benefits are not counted in our official income or poverty statistics. Omitting those leads to a huge understatement of the resources available to those at the bottom of the distribution.
So if we were sitting here right now, and we had someone who’s an inequality researcher who focuses a lot on income inequality and does not focus on the consumption numbers, what would they say? Because if you read about these stories in The New York Times, you hear far more about the inequality of income than consumption.
So, economists are pretty much united that, in principle, consumption is a much better way to measure people’s well-being, and the resources that are available to them, than income. And that’s because income varies quite a bit over time when people take time off to go to school, when they are between jobs — it varies a lot for transitory reasons, while consumption is more of a long-term measure of a household’s well-being. It captures the fact that people are able to save for a rainy day. When properly measured, it accounts for the fact that if you already own a house, you own a car or two, you get a flow of resources from owning those.
So, you can take the case of a retired couple: They may not have much in a way of income, but they may be drawing down their savings, they may already own a car or two. So, they don’t have to spend much to live at a very high level, and they don’t have to have an income because they are able to draw down what they’ve earned over their lifetime.
We’ve been talking about the top 1%, but as you mentioned at the beginning, a lot of your work is not looking at the 1% versus 99%, it’s looking at the other ratios: the 10% versus the 90%. What are your findings there, and what is the relevance? Why is that ratio important and what does it seem to mean?
Sure, if you look at the ratio of the income of the 90th percentile to the 10th percentile, or do the same thing for consumption, that gives you a good idea of the spread of the broader distribution, where the bulk of people are. And since the 60s, there has been about a 30% increase in the ratio of the 90th to the 10th, if you look at income. If you look at the consumption, it’s much lower, it’s only about 7%, and almost all of that increase, essentially all, is in the top half of the distribution — the ratio of the 90th to the 50th. If you look at the ratio of the 50th to the 10th, it’s been basically flat since the 60s, though in recent years it’s gone down as we have increased the transfers to those at the bottom.
What are some of the factors driving those numbers at the 90th and 10th level?
So, the big problem in using income to look at well-being at the bottom, say around the 10th percentile, is that people are increasingly unwilling to respond to surveys and to respond accurately to them. This is a pattern that a lot of statisticians and social scientists have noticed, and what I have worked on in my research with co-authors Wallace Mok, Jim Sullivan, Nikolas Mittag, and others is that the reporting of very important sources of income and well-being at the bottom is increasingly poor. So only about half of the people that are getting food stamps, only about half of people that are getting pensions or cash welfare, report it in our standard income survey. If you’re missing these important sources of income, it makes a lot of people who maybe aren’t middle class, but certainly aren’t starving, look like they are in very bad circumstances.
Correct me if I have the numbers wrong, but if you look at just the official poverty rate measure you hear about in the news, since 1980 it seems kind of flat. But if you look at consumption, poverty has gone down fairly considerably. Is that right?
Yes, so one of the statistics that I like least and I think is most misleading is the poverty rate. The official poverty rate says that we are at the same level of poverty now as we were in the 70s, which just does not fit.
So the Great Society failed, we spent all this money and poverty is no better off.
That’s essentially what the official statistics say, but you shouldn’t believe them for two main reasons.
First, the official statistics don’t count much of what we’ve done to reduce poverty; so the official statistics look at pre-tax money income which omits the earned income tax credit, which omits food stamps, it omits housing benefits, it omits Medicaid. So, it gives you a very distorted picture of how those at the bottom are doing.
The second big reason that the official poverty statistic completely misleads the people taking them at face value is that the thresholds above which you have to be to not be poor go up too fast over time because they are indexed to inflation in a way that overstates the effects of inflation.
And you can see that again if you look at material circumstances in more objective ways of those at the bottom. If you look at the housing conditions of the bottom 20% of the income distribution, they look like the housing conditions of the middle class 30 years ago. So, the rates of air conditioning, central air conditioning, of washers and dryers in the apartment, have gone way up. The incidents of peeling paint, of water leaks in the ceiling or in the pipes, and the like have gone way down.
You’re describing my house. When I think about the house I grew up in, it was about half the size of my current house, it had no air conditioning, and I remember the leaks in the ceilings. I certainly am sure that my parents considered us a solidly middle-class household. But I guarantee if I drove by that house with my kids, they would say, “We didn’t know you were so poor.”
I think that’s exactly right. I think that’s what a lot of us can see in how our lives have changed, but the official statistics don’t really reflect that, in significant part because of the overstatement of inflation and because of the omission of in-kind transfers and other government benefits.
I want to talk a bit more about poverty. What do you feel comfortable saying about inequality in America in 2017? The headline number we see, that it has gotten a lot worse and it is hurting the middle-class and mobility, and therefore is the key challenge for policymakers — is that more or less correct?
Well, my research says that inequality — at least in the bottom half of the distribution — hasn’t gone up over the last 30 years, and it’s even gone down a little bit in the last 10 years. So, since I’m mostly concerned about inequality at the bottom rather than the very rich —
Well, I always know someone’s going to be richer than I am, and I just don’t fixate on that much. I’m more worried about those at the bottom who may not have enough resources to partake in society and reap some of the benefits of economic growth. And, what I think you see is that the bottom has been pulled up as the economy has grown over the last 30 years. And, we have talked about how you can see that in tangible terms in people’s housing. You can also look at the characteristics of cars people drive — your inexpensive car now has safety features that you only got in very fancy cars 20 years ago. And, its performance in terms of acceleration and gas mileage is so much better than cars 20 years ago.
You can also see this looking at people’s food-consumption patterns. It wasn’t that long ago when we were worried about people not getting enough calories; now, we’re more worried about getting the right calories and issues of obesity — it’s just a very different conversation than it was 30 or 40 years ago.
Maybe some people don’t value this, but I certainly value the fact that now my kids have access to all the world’s information via computers. I certainly didn’t. I did not have a computer in my home, so that meant maybe once a week I would go to a library. I mean, that’s huge. Do the income statistics take that into account?
I think we have a very hard time accounting for new goods and for goods that take a lot of people’s time. And people devote a lot of their leisure time or even their work time to using the internet or smartphones or computers and iPads. We have a hard time accounting for improvements in those types of goods. Especially when people are spending a huge amount of their time using them, even though the goods are essentially free — the internet is certainly free once you’re on.
So I think you’ve hit on something, that the ability of a middle or low income household to get access to information, to get access to entertainment — from music, TVs, movies — it’s so much easier now at low cost. And that’s a huge improvement in quality of life that stretches across the income distribution.
We have a lot of a common numbers most people see in the media, whether it’s poverty or GDP or the unemployment rate. Are they all broken, or built for a different era? This is important because these numbers drive all the news media coverage that drives a lot of our political debate. Do we have, as economist Joel Mokyr at Northwestern University has said, economic statistics built for a wheat and steel economy while we are in a digital age?
I think that’s exactly right. We also have, besides these changes that have meant that there are all these new goods that are hard to value — from Facebook to Twitter to texting — we also have this problem that people don’t want to respond to surveys anymore.
How does that vary by income level?
Well, it really is across the board, so it’s not something that disproportionately affects low-income or high-income people. Everyone these days gets a constant barrage of surveys. If you’re at home, answering the phone, you’re going to get surveyed about everything from what radio station you listen to to political issues, and if you buy something online you’re going to get a survey about that. If you go see a doctor, you’re going to get surveyed.
People are overwhelmed by these surveys and are less willing to respond, and that’s true across the income distribution, and it means that our statistical agencies have a very hard time doing what they used to do. It’s just gotten harder and harder because people are less cooperative.
And therefore policymakers are flying blind in some of these areas.
You mentioned surveys. I think if you surveyed a lot of our economists, they would tell you they like the Earned Income Tax Credit as an anti-poverty policy. Is this an accurate assumption?
Yes, they like it because it transfers income to those who are needy in a way that does not discourage work; in fact it tends to encourage work. It’s well targeted because it goes to people who are low income, particularly those who have children. It’s bigger for larger families that might have more of a need for help. So it really has a lot of attractive features.
So, what about that fraud issue many conservatives gripe about? Is it an overpayment or mispayment issue? What is happening there with that?
I think that that’s overstated. You sometimes will hear numbers like 30% — I probably shouldn’t repeat these numbers — that a substantial share of the dollars are paid in error. But many of these calculations don’t go back and say well, “The money went to grandma instead of her daughter who we think should get the credit.” Those statistics often will count the money that goes to grandma but then not net out the money that should have gone to the daughter, so it would’ve been paid out anyways; it just went to the wrong person. Or, if a family really should’ve claimed three children rather than two children, they’ll call the entire payment fraud while really there’s an error and it’s just part of the payment.
Now, you don’t also want to call these errors fraud. It’s a scandal that the instructions for the Earned Income Tax Credit — I haven’t looked at the most recent tax guide — but it used to look something like 18 very dense pages out of the entire 200 pages of tax guide, so it’s just way too complicated. It’s not surprising that there are lot of errors under that circumstance.
One good thing about the EITC is if you like bipartisanship you can find people across the political spectrum who think it’s a good idea. Some people think it’s such a fantastic idea that we should be doing a lot more of it. Ro Khanna, a congressman from Silicon Valley, said he would like to expand the EITC by over a trillion dollars over 10 years to respond to the challenge of automation and stagnating wages. So he wants to expand this as a massive work support program for the new economy. What do you think about that idea?
Okay, I don’t know the details of this, and I would really have to look at it, but I would be concerned if these are payments that go to everyone regardless of work. I’m skeptical that we could expand the EITC a lot more from its current level, except for maybe the childless. But, I don’t see how you get to a trillion dollars.
Some people say, “Listen, if the point of this is to boost living standards of families, why not just double it?” What would be the impact, if we could find the money to pay for it, to just increase it by 50% or 75%? Is there a downside to that?
Well, there’s no downside if you have the money come out of the air, but in order to get that money, you’re going to have to raise taxes someplace else.
But would it affect the recipients’ incentives in any way, or their behaviour in a way that we wouldn’t want?
I don’t think so. If you could raise the return to working for those at the bottom, that would be a great thing. We have too many people who currently don’t have jobs. Work is something that is ennobling, it is something that encourages those around you to do if you do it, and it’s something that if we can subsidize in a sensible way, I think subsidizing work is a good thing.
What do you think about the idea of a Universal Basic Income, especially if we have a changing economy with very uncertain effects on jobs?
Let me first say that we have a small version of that. It’s called SNAP, and we have almost 50 million people getting SNAP now. But thinking about a much bigger Universal Basic Income I think is not really a good idea. There’s a paper from about 30 years ago by the economist George Akerlof, a Nobel Prize winner, and I thought this very simple argument was pretty convincing. If you give a basic income to everyone that’s really enough to get by on, then the tax rates for everyone else are just too high.
Our system — which targets those who are really needy for assistance, those who are disabled, are aged, those who cannot support themselves — that’s much more sensible, and that allows you to provide a sufficient amount for those who are needy without taxing everyone else at a higher rate that would discourage their work.
Comments are closed.
1789 Massachusetts Avenue, NW, Washington, DC 20036
© 2018 American Enterprise Institute