The Impact of Government Programs on Wealth Inequality

I want you to think a little bit when we're thinking about kind of societal optimum, right? You could think of two hypothetical extremes. One is, no government, no taxes, no spending. The other one is the government has 100% share in the economy, a 100% income tax rate.

So what would these two things look like? I mean, maybe in practice, the two kind of examples that are closest to that are, say, the Cayman Islands. On the one hand, where there's no direct taxation, revenues are generated from maybe some import duties and some financial service fees, some work permits.

There's basically no taxation, no direct taxation, and as a result, there's basically nothing the way of government services. The Cayman Islands are primarily just a locus for being a tax shelter, that's what they are. Or you could think of, say, a country like Cuba, where there is essentially almost no private sector, the private sector only comprises about 12% of employment.

So it's second to North Korea in its lack of economic freedom, and implicitly, that's just a very, very high degree of taxation. So there you have a situation where the government basically is imposing very, very high taxes on the economy. And that's contributing to the fact that there's just not that much economic activity in Cuba.

Because think about it, why would you work or start a business if the government's gonna take 100% of your earnings? There's just no incentive to do it, right? So these are two extremes, I don't think very many people in this room want either of the extremes. And the question is really gonna be, where in the middle are we gonna land.

So let's talk about what the academic world has done to contribute to this debate about income inequality. The media focus on US inequality, most of those articles that I showed you on the previous slide focuses around three economists, three French economists, Piketty, Saez, and Zucman. So Thomas Piketty, French economist who's a professor in France, Manuel Saez, who's a French American professor at UC Berkeley, and Gabriel Zucman, who's a French professor at UC Berkeley.

And they present income and wealth inequality as a very severe issue in America, largely because they show statistics that demonstrate increasing pre tax income inequality. And there's been a lot of media coverage of them. And what we're gonna do in this talk is we're just gonna look a little bit at what the evidence is, what the evidence is that they're showing, and what the overall evidence is.

Cuz it turns out they're not the only researchers who have studied this question, but they are the researchers who get the prevailing attention of the media, of the news media. So I asked the French revolution, because these researchers also kinda go out there pretty far, and they say they make statements like that the top marginal tax rate in the United States should be 75%.

Or the optimal marginal tax rate, top tax rate should be between 72 and 82%. So those are pretty high numbers, all right? The top federal marginal tax rate right now is 39%, you add on state and local and things like that, also in California, you have high rates where maybe around 50%.

You add on some self employment taxes even higher, but they want it to go significantly higher. So question is, are we, should we have the French revolution as the result of their findings? So let's have a look. I resisted building in a playing of the Marseillaise into this presentation.

I thought a little music would be good, but I usually have to rein myself in with a humor, as you saw last night during my opening remarks. Okay, so let's look at the Piketty Saez approach. And I'm using the French flag as a little symbol because I don't wanna keep saying Piketty Saez,Piketty Saez, Piketty Saez, Zucman.

These are just names, they may be the first time you've ever heard these names. So it's sort of a mnemonic to think, okay, this is maybe these are the French economists, okay, and so let's look at what they show. So they, starting in 2003, they started going back historically and tracking top income shares.

And so here's the top 10% share of income that they presented, it looks like it started around 32%. By 2020, it was around 48%, so they're saying the top 10% are earning 48% of the income. Here is the top 1% share according to them, right? This is actually quite similar to what we saw in the poll that we did, right?

It started out around 10% in 1960, it was fairly flat through 1980, between 1980, and 2000, it seems like it went up quite a bit, and then it kinda flattened out over the last 20 years. It's actually been quite flat over the last 20 years. These graphs are going to 2019, they've also extended these through 2021, and they're working on doing it beyond.

But because there are a number of extensions of what they've done that are not available for that date, we're gonna look through 2019 in this presentation. For 2021, 2022, they're projecting that this 1% number was around 22%, like what we saw. So if you're looking at the details, that's something you might notice.

Okay, so they found vast increase in US income inequality starting in the mid 20th century, at least that's how they put it. I would say most of what they found, just looking at their data, is 1980 to 2000, that's kind of when most of the action happened. They're using data from tax returns, so they're presenting pre tax income.

So this means pre tax and also before transfers from the government, so when the government taxes people, some of that money is being used to provide government services. Some of it is also being used to provide income support, or welfare, or medical services to lower income people. And so none of that is considered in this series, this is basically just how much income did you earn from your work, okay?

And that has clearly gone up, I don't think there's any doubt about that. That income inequality based on how much you've earned from your work market income that is, increased, most of it from 1980 to 2000. What does it exclude? It excludes Social Security, so Social Security is our income support program for the elderly, and it's quite progressive at middle income levels.

Transfer payments, so any payments to individuals through things like, food stuffs, or welfare programs. Employer healthcare contributions is another one, if you work for an employer, they contribute pretty substantial amount of money to your health insurance, that's not in there, all right? So there are a number of things that aren't in there.

And these increases that they've documented have been highly influential. Peter Orzegae, former director of the Office of Management and Budget of the federal government of the United States the White House in 2009 said. Saez helped shape my own thinking and had no small influence on President Obama's budget.

So Piketty and Saez updated their data a little bit in 2018, they updated their approach in 2018. To some extent, this is what we saw in the last slide, just a repeat. They did a couple of additional things, first of all, they revised their numbers to some extent to reflect what they call 100% national income.

So now they're moving away from that AGI measure, moving away from that adjusted gross income measure. They're trying to include other income items that don't necessarily show up on the tax returns, they're trying to include 100% national income. So that includes transfers through the US government's various welfare programs, through their healthcare support, through their various income support programs.

through food stamps, through women and children, through all these various programs that the us government has. And what have they found? So, first of all, their pre tax numbers are now a little bit different, it's somewhat flatter. Still an increase, but somewhat flatter, and they've also provided a set of after-tax numbers.

So after-tax, it looks like what we've seen, is we've seen after-tax income inequality of the top 10% share started about 31% in 1960. That's the top 10% share of income, and it's risen to about 40% today. So they did do that, including these after-tax, after-transfer series. And that's important, because it's important to reflect the fact that over this period of time.

The US government, particularly the federal government, but also state, local governments, have been doing a lot more to actually redistribute income. To tax higher-income people and to redistribute some of those taxation revenues to lower-income people. They still find that the top 10% received around 55% of the economic growth from 1979 to 2014, and they're very on the French solution.

The US must adopt a more strongly progressive tax system, they want a more strongly progressive tax system. You voted on what you thought the tax system should be. It seems like the majority of you, when given the facts, don't necessarily think we should have a more strongly progressive federal income tax system, but they're very much pushing towards that.

Redistribution, we're already doing a lot, here are some other measures of redistribution that might be of interest. The left graph shows the federal government only and the right graph shows federal, state, and local. So this is showing average transfer income versus average federal taxes paid. Transfer income means income that people receive from government programs, income people receive from government programs.

So starting at the lowest quintile, the lowest quintile of individuals are receiving income from government programs that average around $18,000 per year. They're not paying anything significant in federal taxes. The second quintile, less transfers, paying some taxes, you can go up the quintiles. In the fourth quintile, they're paying around $20,000 per year in federal taxes, getting a little bit of transfers.

And the highest quintile are paying around $79,000 on average in taxes and not receiving transfers from the federal government. Right graph now starts looking at federal, state, and local sources combined, because we have multiple levels of government here. Local government, state government do a lot of things. A lot of our education is provided through local governments.

A lot of our public safety is provided through state and local governments. And what we see here, these are what are called fiscal incidence rates, which you kind of think of as tax rates, net of transfer rates. So what are you having to pay in net of what you're getting out?

And once you add in the state and local, it is a little bit less progressive than the federal level alone. But you do see here taxes, so bottom quintile is paying around 25% of the taxes, and top quintile, 35% of taxes. So there is more progressivity in the state and local side that's adding to that.

This is what happens when you combine the tax rate with the fiscal incidence rate, how much transfers they're getting. And you see that the bottom quintile, net of the transfers they're getting out, have a net fiscal incidence rate of around 10%. And the top quintile have a net tax rate, net fiscal incidence rates around 41%.

So we are actually already doing quite a bit of redistribution in the United States.

So this is kind of a summary of where we are so far. This shows these Piketty and Saez graphs superimposed all the lines here. So the blue lines are the original Piketty and Saez series, the French series, I might call it, just for ease of referring to it.

The red line is their revised series, once they took into account all national income or tried to account for these other income sources. And the green line is what happens after they accounted for the taxes and transfers. Now, what happened then in the debate? Okay, enter the government economists, so these folks are not as flashy as the French economists.

They work at places like the United States Department of Treasury and the Joint Committee on Taxation of the Congress of the United States. They have very, very demanding day jobs. I spent a few months in Washington, DC, at the Council of Economic Advisors. I met some of the people who are on these committees, working in these agencies, they don't have a lot of time to just sort of pursue research on their own, and say, what would be a useful thing to look at?

They are busy people. In fact, one of these authors that I'm gonna talk about now, a guy named David Splinter. He's a research economist at the Joint Committee on Taxation in the US Congress since 2012. These are nonpartisan organizations, by the way. He writes on his CV, David Splinter, that he has scored the tax impacts of thousands of pieces of proposed legislation in the US Congress.

So my point is, he doesn't have that much time to just be sitting around checking up on what economists at Berkeley are saying about income inequality. Yet he and Gerald Auten, who's a research economist at the Treasury since 1987, saw this Piketty and Saez work. They saw the work coming from the French economists, and they said, something is not right here, because these people have worked.

They have access to the statistics directly through, direct access via the US Treasury, and they came up with some points that are problematic. So in fact, what they did was they provided some additional, some new income series, top 10% and top 1% share series. That adjust for a number of things that they find are problematic that Piketty and Saez do not consider.

So some of these can get a little bit technical, but let me tell you about what some of the main ones are. So one is family size. So compared to, say, 1960 or 1970, family sizes are on average somewhat smaller. And they're smaller for lower-income people more so than for higher-income people.

So there's just a lot more people who are single, essentially, as opposed to married couples. And so if you're not considering the fact that that family structure has changed, you might misattribute, we're looking at tax filing units, household units. So we might not properly reflect what's going on with income inequality.

So that's one issue. Another issue is dependent tax filers. There was a major change in tax law in the 1980s that has continued to have an increasing impact over the decades that followed. Which say that if you're a dependent on someone's tax return, which many of you might still be dependent on your parents' tax return, but you're earning money, you have to file a separate tax return.

So those people are not, they're being counted in Piketty and Saez as just, okay, these are poor people, right? You might have only earned a few thousand dollars this year as part of a summer job, but you're also dependent on your parents. So since the 80s, those dependent tax files have grown a lot, and that adds a lot of lower-income people in who aren't really lower-income people.

So that's another issue that the government economists took some issue with. And finally, allocating underreported income according to IRS audit studies. So one of the points I made about Piketty and Saez is that they use tax data, they use what they can get from tax sources to do their work.

And one of the problems you might say with that is, okay, well, what about just unreported income from people who are just not reporting their income accurately, people who are just basically evading taxes? And Piketty and Saez, they allocate underreported income in one way that's equal across the income distribution.

But Auten and Splinter actually allocate unreported income based on what they know, which is based on IRS audits. So when the IRS goes and checks up on whether you paid your taxes or not, they go and they find this out. And it turns out that while in dollar amounts, higher-income people are the sort of fruits of their tax evasion are higher than that of lower-income people in a percentage way.

Actually, percentage wise, the lower income that's reported, the more evasion there actually is. So the government economists came in and they actually found much less income inequality. To the extent that, actually, if you look at the bottom line of these two graphs here, This is after tax, after transfer income inequality once they adjust for some of these changes, it's a flat line.

The press doesn't report this at all, you don't hear about this at all. These are government economists coming in saying university economists are not properly using government data but that is not reported in the media at all. And that is a big problem because in the media there is a strong narrative of exploding inequality.

I'm going to come back to some aspects towards the end that I think are very valid in what pike and sayes are saying in their work. Particularly what I think is very valid in their work is the fact that pre tax, pre transfer income, it is the case that there is more income inequality now than there used to be.

Something is going on with their returns to skill, to high skilled work. I think that the narrative that comes around that, we have to do more taxing and more redistribution. I don't know because after tax after transfer, it's been pretty flat, according to the treasury and Joint Committee on Taxation economists, doesn't get reported in the media so that's a problem, okay.

And I focus a lot on the top 1%, it's not just the top 1%. This is real median household income in the United States. The median household income has grown substantially. This is inflation adjusted 55,000 in 1984 to 70,700 in 2021. A right graph shows real after tax, after transfer income across quintiles from a study by Elwell et al and they show that the first quintile actually has seen the most growth in after tax, after transfer income since 1959 because there's been a big increase in social spending.

Other quintiles have seen pretty similar around the middle, 160% or so in the top has seen about 175%. One thing I'll also come back to is the question of how much you should we really care about how different these are across the quintiles of income, I'll come back to that.

I want to talk for a few minutes about wealth inequality, so income disparities have not risen that much after tax what about wealth disparities? So let's examine a few sources, okay. One is Saez and Zucman since their unifying theme without Piketty is that they're professors at the University of California at Berkeley.

I'm going to just start calling these the Berkeley estimates, if that's okay. Cal estimates, that is okay. Just it's easier to do that than to always refer to their names over and over again and there have been three sources one is their initial estimates. One is their estimates with some revised technical assumptions, and the other is some revised technical assumptions and fully considering all wealth sources, including Social Security.

One of the challenges with this is that the United States does not collect any administrative data on wealth. No one has to report their wealth to the government, except I, actually, when they die, if you die you have an estate that you then have to declare. Your heirs have to declare the value of your estate because we have estate taxes but there's not a centralized source for wealth, so that makes this wealth inequality very difficult.

So Saez and Zucman, the Cal folks, they claim massive increases in wealth concentration over the past five decades. In fact, this graph looks a lot like their income graph, and that's kind of not surprising, because the source for building this is actually the same source as they're using for their income inequality measures.

What they're doing is they're taking those income measures on tax returns, and they're trying to sort of back out from that what people's wealth must be a tricky exercise. So, for example, suppose someone reports that they earned $1,000 in interest on their tax return. Well, was that at a 1% interest rate or at a 10% interest rate?

If you choose a rate that's too low, you're going to overestimate the amount of people's assets. If you choose a rate that's too high, you're going to underestimate the amount of their assets. Another issue is that it excludes the fact that we have the Social Security program, which is for households actually a form of savings.

So other authors have revisited this, one set of authors, reevaluated the technical assumptions, didn't change it that much. They basically dug into these interest rate assumptions, they found that they were a little off. They still found that there was high inequality, albeit a little less than the Cal researchers.

This is saying the top 1% have around 34% of the wealth, so that's more inequality than there is in income. Other thing we have to add to this is the Social Security idea, so, Social Security is the near universal retirement income program in the United States every employee pays in 6.2% Social Security.

Employers pay in 6.2% as well, both up to a taxable maximum of $160,000 per year. In exchange, you, when you retire, are entitled to a monthly income check the more you pay, the more you get in. As you read in the press, this program has some financial challenges but certainly not gonna be the case that there's going to be no Social Security annuity when you get to retirement.

Worse, people like you, you might find yourself receiving 20% less than what you would without these fiscal problems. But Social Security is a kind of wealth, and this has been reflected in a lot of research that came after Piketty and Saez as well, how does this work? Put yourself in the shoes of this young individual on the left of the graph.

This person, when they start working, they begin paying a portion of their income to the government for Social Security that continues until retirement. What is this person thinking, this person is thinking, well, the government has taxed me, but they owe me money in retirement. I'm going to get an annuity when I retire, I'm going to get money from the government.

And from the perspective of the person who is retired, they're receiving this money and they're going to keep getting it. I've been getting money from the government since I retired, and I will keep getting it. So for those of you who are economics major economics people you might recognize the concept of present value when you know that you're going to get a certain income stream, or even if that income stream is a little bit uncertain, but you know you're gonna get it.

That is like an asset that is valuable, and that asset is very, very valuable to middle and lower income people in the United States. So some other researchers have gone ahead and they've said, okay, we're going to incorporate that Social Security as wealth. And when you include the value of social insurance programs, especially Social Security, you actually find that wealth inequality in the United States hasn't really moved.

They also do adjust for the possibility there might be benefit cuts or tax hikes, and they still find minimal growth at the top 1%, 10% because once again, similar story to the income inequality. There's a government program going on in the background, and that is providing people with significant resources.