Research perspectives on the costs and benefits of a hot economy
View the presentation slides (pdf, 2mb)
Welcome to this first panel of the day. The goal of this panel was really to dig a little bit further on what the economic research had to say about the costs and benefits of running a hot economy.
I am joined on stage with Doug Holtz-Eakin, Heather Boushey, Jason Furman, and Sarah Bohn. The rule of engagement here is that we're giving them 10-15 minutes to present their perspectives, and then we'll open it up to a Q&A session. The first speaker will be Jason Furman, who is currently a professor of practice of economic policy at Harvard University. But before that was chairman of the Council of Economic Advisers from 2013 to 2017. We're delighted to have you here.
Thanks so much for including me. Thanks for this really important discussion. The only thing I have to add to what Vice Chair Rich Clarida's excellent remarks and my Dean, Doug Elmendorf's excellent remarks, is a bunch of slides to make a lot of the same points that the two of them make.
I should also apologize in advance. I didn't have as much time to prepare this that I'd intended to have. The session was about the cost and benefits of a hot economy. I only had enough time to prepare the benefits, so I'll just tell you that at the outset and refer you to Doug's highly caveated remarks for the costs.
Okay. Let's start by looking at where we are now in terms of unemployment rates and how they compare to where the economy was in the quarter before we went into the Great Recession, pretty much the peak of the economy before the global financial crisis.
Just about every measure of unemployment is lower today than where it was before the crisis. You see that in the overall unemployment rate, which is 1.2 percentage points down. Of broader measure, U6 actually incorporates people that have given up looking for jobs that would take a job but aren't actively looking for one, and people that are involuntarily part-time. That's fallen a little bit more to an extraordinary 7.2%.
You've seen, as you generally do or almost always do, larger declines in black unemployment and Hispanic unemployment. You also saw much larger increases for both black and Hispanic unemployment rates in the recession. One of the benefits of a hot economy is that it brings that down. And longterm unemployment had actually for a long time been on an upward trend for several decades. I didn't think we'd ever see it below 1% again. These are people unemployed for 26 weeks or longer, but it's actually fallen now to 0.8%.
Now let's shift though and look. Unemployment rates are just the people who say they're looking for work. Now let's look at everyone, the entire population. These are now non-unemployment rates. This is the opposite of what's often called the employment population, right? And there we see a more mixed story. A larger fraction of people 16 to 24 aren't working. That's a little complicated because more of them are in school and more of them while they're in school aren't working. It's a topic we could talk about the pros and cons of another time.
Older workers are actually working at a higher rate. The prime employment rate as of Q2 is back to where it was, and if you look in August it was actually improved from where it was. But for men you have an employment rate that's nearly a percentage point higher than what it was. Women, you have an employment rate that's much lower than it is for men. There's a whole set of issues around paid leave and the like we could talk about, but I want to focus some on the more cyclical issues now and zoom in on those prime age men.
The employment rate is a function of these two. The first is the unemployment rate, and the unemployment rate for prime age men rose in the recession and came down a lot since then. So the unemployment rate for prime age men is down. The problem is their labor force participation rate is also down. A lot of these men have given up looking for work and have left the workforce. These are mostly men with a high school degree or less. They're mostly not married to a woman who is working or have somebody else in the household who is earning money. And as a result when you combine these two, you see that even though their employment rates have improved, it still hasn't gotten all the way back down to that dotted line of where it began.
This isn't just some unique thing that happened in the wake of the global financial crisis. If you look at men age 25 to 54, their non-employment has been on an upward ratchet for decades. Since the 1960s, every time you have a recession, the non-employment rate goes up. Every time you have a recovery it goes down, but it never gets back to where it started. It always starts the next recession from a worse place than where it started the previous recession, and keeps going up and up.
If you look though that red dot for where we are now is a bit below that trend line. It's frankly come down more than I would have predicted and more than I was on the record saying a couple years ago. So in terms of two lessons from this, one is monetary policy may have more scope than many appreciated. There's been a really big improvement in the non-employment rate for prime age men. There may well still be more room for that to grow even without the unemployment rate moving by bringing more people into the workforce.
The other lesson from that chart is that there's a limit to what the Fed can do. This is 10 percentage points worse than what it was in the 1950s. Monetary policy maybe could do one percentage point of that, maybe even in some amazing world, two percentage points of that. The bulk of it is due to other factors: education in support of labor markets, limits on mobility, lack of training programs, mass incarceration, opioid epidemic and more. You know, what we sometimes call structural policies. And so you need a combination of both of these types of policies, some that the Fed can do, some that they can't.
If you're asking about the benefits side of a hot economy, the employment increase by itself is a huge benefit. In many ways, the type of inequality I'm most concerned with in the economy is people who are outside of the economy entirely and who don't have a job and face the type of social exclusion and lack of purpose. In many communities, that is the case.
There's another question of what does a hot labor market mean for wages? In particular, and we heard about this in the context of the Phillips curve, it means prices go up more, but it also means that wages go up more. So it's an empirical question as to which of these go up more. Surprisingly, fewer people have looked at this than you would think, and when I started asking this question, I got almost nervous that what if you run a hot economy and you run a hot economy, it means prices go up more than wages go up, and so you end up with real wages being lower? Now, you'd still get the employment effects. We know the employment effects are large. We know the employment effects may be even more important than any of these wage ones, but what's going on in wages?
First of all, just to look at some summary data on the left there, there's a whole bunch of different measures of wages. I'm focusing again on age 25 to 54 because they take out a lot of the demographic things that affect these wages. You see nominal wages have been on an upswing since about 2010. On the right, you see real wage growth. I'm using Core PCE to effectively extract a trend so it's not bouncing around a lot as global oil markets bounce around. You're really looking at the US economy. You've seen an upward trend in real wages.
One thing to notice, and I'll come back to that, even with this upward trend, even with this hot labor market, real wage growth is considerably below where it was in the late 1990s. It was around 4% then. It's considerably below that now.
These are of course averages. They look at everyone. As Doug talked about, if you drill down and look, the largest wage growth has actually been in the bottom quintile, faster than the wage growth for any of the other quintiles. And in fact, even though overall wage growth isn't as good as it was in the late 1990s, if you're in the bottom 20% of the income distribution, you'd rather be in today's labor market than that labor market. In fact the bottom 40% that's generally true for. So how the wages are distributed matter a lot.
We were talking about the shapes of wage and price Phillips curves. Just to show you that, on the left-hand side, this is the relationship between the unemployment rate and average hourly earnings growth. This is nominal. On the right-hand side you see the same exact thing, but here it's for core prices.
The price one is very flat. The wage one is much steeper, which is to say if you lower your unemployment rate by a percentage point, it appears to raise your wage growth more than it raises your price growth. In fact, you have to squint to see what it does for price growth.
These are of course just two particular ways of looking at it. I've been working on something, and it turns out this is quite robust. These are hopefully the only regression coefficients anyone will show today, but the point of this is is there's different ways to do the Phillips curve. We heard about that. A traditional one, an accelerationist Phillips curve, which looks at the change in the inflation rate, or an auto-regressive Phillips curve, which looks at the change in the inflation rate, but models it differently.
There's different measures of slack, and what's interesting here is I used a measure of slack, the unemployment rate. I used the short-term unemployment rate, which is a very narrow definition of slack, and I use the prime age employment rate, which is a very broad definition of slack. What you should take away from this is in every one of these pairs, the coefficient on average hourly earnings, that's how much an improvement in the labor market helps earnings, is larger than the coefficient on prices, how much a hot labor market raises prices. It's true with these three Phillips curves, these three measures, and I deleted the other eight slides I had that did this for different measures of prices, wages as well.
Now this is 2000 to the present. If you look earlier periods, you don't see anything like this really robust relationship with wages, this lack of one with prices, and this is the period when inflation targeting has been very credible. Basically had the same inflation rate the whole time, give or take, and that means you're not going to see a Phillips curve effect in prices. Effectively, the Fed has undone that and has what remains is a Phillips curve in wages, not prices.
I said wage growth isn't as high as it was in the late 1990s. That's because productivity growth isn't as high as it was in the 1990s. The Fed could maybe help with productivity growth, but not that much. So once again, there's a whole bunch of other things that even if you're running a hot economy, if you're running it in a world of 1.2% productivity growth, there's a limit to what you can accomplish compared to the late 1990s when you were running a hot economy with 2.8 productivity growth.
So in summary, when you look at wages, I think you get the same story as when you look at employment rates. Monetary policy does appear to be able to raise real wage growth because it raises wages more than it raises prices. But monetary policy by itself can't offset decades of structural problems. It can't undo all the different causes of higher inequality, it can't undo all the different causes of slower productivity growth. And unless we do that, we're not going to be able to get all the way to the type of wage growth that we want to have. Thank you.
Thank you, Jason. To start us off on the panel, the next speaker will be Heather Boushey. She's the founder and executive director of the Washington Center for Equitable Growth. And she's also an author of a forthcoming book on income inequality that's going to be published by the Harvard University Press. It's going to be coming out next month, I believe.
Next month, yes. Very exciting. Thank you.
When they asked us to put this panel together, they asked us to not repeat each other too much. So I tried to come up with some different things to say than I thought Jason or other folks on the panel would say. But as we've had this conversation this morning about the benefits of the hot economy, we tend to start with the aggregate benefits. And Jason talked about this and Doug and everyone else as well, that we talk about the aggregates. Low unemployment leads to stronger wage growth, and how we see that this benefits especially folks at the bottom the most, that we're seeing that as a hot economy moves forward, it reaches disadvantaged groups.
I wanted to just start off by giving a quote from the recent Jackson Hole speech by Fed Chair Powell, who emphasized how the high pressure job market was reaching people in "low and middle-income communities" and leading employers to "train workers who lack required skills, adapting jobs to the needs of employees with family responsibilities and offering second chances to people who need one."
So there are a lot of benefits, and as Jason really focused on the good side. But I want to spend a few minutes talking about how that's not enough. One of the things that we see, and I think that we have to, as we're thinking about what economics can tell us about the benefits from a hot economy, we need to understand that we're starting from a place of an economy with extremely high inequality and the likes at which we haven't seen in earlier eras that affects how we think about our models.
As I was listening to Doug talk, I was sort of sitting there thinking about what an analogy to that could be. Often now as we're getting into fall, I start questioning whether or not I should really be buying myself a very thick investment kind of winter coat. And each year I don't because I'm not actually sure that in DC I'm going to get enough wear out of it as the climate changes. And maybe it's a very dark analogy, but I feel like there's an analogy here to how we should be thinking about both monetary policy and fiscal policy more generally with respect to inequality. That the world has changed markedly in terms of the foundation upon which we're doing policy.
So I wanted to spend a few moments talking about some of these structural effects of inequality. I want to encourage us as we're thinking about the benefits of a hot economy or monetary policy more generally is to think about both the effects of policy on the distribution, but just as importantly the effects of the change distribution on the effectiveness of our policies as well. And I know that there are folks here in the audience that have spent a lot of time thinking about that and hope we can spend some time talking about this today.
Just a quick note, that is of course one of the goals of my center, the Washington Center for Equitable Growth. We've given away over $5 million to over 200 scholars to investigate whether and how inequality affects economic outcomes. One of the things that we've learned is that there is this large group of new scholarship where people are using new data, data that allows us to look at heterogeneity to understand economic outcomes. I want to speak to a little bit of what we've learned here, that especially this proliferation of micro level data to understand macroeconomic outcomes is an important avenue that we should be considering.
But first, what I like to think is sort of the grim—I mean, unless you like inequality—slide deck. Just a few charts to kind of ground this conversation. Usually I do these as two sides. I'm experimenting to see how this works in audience putting them on one side, but what this is is the data from Piketty, Saez, and Zucman. The X axis is income distribution, incomes from low to high. The top chart looks at income growth over the 1960s and 70s across the income distribution.
What you can see is that the economy grew on average about 1.7%. Most people in the United States saw their incomes grow alongside the average of income growth. So about 67% of people saw their growth grow within about 10% of the average. If you were at the low end of the income distribution, your incomes grew faster. If you were at the top, your incomes grew slower. In the 60s and 70s we were a nation that grew together. Since then of course the trend is markedly different. Not only is overall income, gross national income, growth slower, but only those above the top 10% are seeing their incomes grow as much as the average.
And it means data like this that really should be pushing us to disaggregate our aggregate statistics when we're thinking about growth, what is it that we actually mean? How is this looking across the income distribution? Because this change is so marked and because it means that the vast majority are not benefiting. Doug talked about this in the Q&A that there's an enormous frustration out there, and you can see that in this chart. We used to be a country that grows together, and now we aren't at all. So what does this mean for how we think about the causes and consequences of policy?
Three more kind of ground, stage setting figures here. One is that I think we can't be talking about the effects especially of monetary policy unless we understand the changes in the distribution of wealth and just how much of the nation's wealth has been coming increasingly congealed among a small group at the top. Thinking about what this means, we know that there is a difference in the marginal propensity to consume and to save across the income distribution. As you have a rising concentration of wealth, that's going to affect the effectiveness of economic policy.
This data, this slide, usually I show the slides the longer time series from Saez and Zucman. It gets confusing which one's done which. But this is actually data I'm really excited about that just came out from the Federal Reserve that is this new data that's integrating the Federal Reserve financial accounts with the data from the survey of consumer finances, which gives us a new sort of more up to date understanding of what wealth looks like.
Penultimately, we have to be thinking as we're thinking about both effects of inequality on and through policy, the rise in market concentration. This is something that Jason actually elevated a lot at the end of the tail end of the Obama administration through the CA, that this is something that we haven't been talking about enough, what this means for labor market outcomes in particular in terms of the effects of labor market monopsony, but how that also affects the effectiveness of monetary and fiscal policy.
So I'm going to skip this one and go straight to this one. One of the things that we know over the preceding economic expansions and contractions is that the richest 10% of Americans have received about half of all growth in recent expansions. Now, Jason showed some charts at the end of his presentation showing that in the last couple of years we've seen wage growth start to increase, and that's been at a slower pace than the late 1990s period of full employment.
But it's important to understand just how much of the gains in income are going to the top and to keep that front and center as we're thinking about these effects of a hot economy that we need to, if we actually care about ensuring that we have a broadly distributed gains in wages and income across the economy, thinking about the structural challenges and pairing those with the right kind of monitoring fiscal policy needs to be front and center.
So for example, as Jason put up that chart on wage gains, I was thinking about how over the past couple of years you've seen these very sharp increases in many places across the United States in the minimum wage. There's actually strong research evidence that that increase in gains in wages at the bottom may be just as much due to the increase in the minimum wage as it is to the full employment. So pairing those together is where you get that big bang for the buck.
I wanted to spend the last couple of minutes here in my remarks just with a couple of data points. Thinking about how a hot economy, some of the benefits and how we can think about it in terms of the new research that we're learning. And since Stephanie was not on the panel, I wanted to make sure that we, and I wasn't sure what Jason would show because I didn't get his slides in advance, I wanted to make sure that we had a couple of slides showing the effects of the low unemployment on different demographic groups and income groups.
This is from a recent paper by Jared Bernstein and Keith Bentele that shows how high pressure labor markets put upward pressure on hourly earnings looking across quintiles. The biggest bang is indeed for those at the bottom, which is something we've talked about. But unless you've seen a slide on it and multiple ones, we're all concerned you won't believe it.
But one thing, to also put up something that explicitly gets at issues around race, which is where this is so important to show that if you look especially at black households at the bottom 40% of the income distribution, that having low unemployment gives you that bigger bang for the buck, especially for those families. Of course, remembering that even as you see this bigger bang for the buck, you have a lot further to go. And it still remains the case even though the black unemployment rate has fallen more than white unemployment, that it's still about double the white rate. So that's a structural issue that even as we can say, "Oh look, it's fallen so much," understanding that ratio still hasn't budged a lot is something that I think that we should be keeping front and center.
So there has been a lot of new and very exciting research over the past few years that has focused on making this explicit connection between monetary policy and inequality. A recent paper by University of Texas at Austin economist Olivier Coibion and his coauthors finds that expansionary monetary policies from the Fed both lowers income and consumption inequality across households while contractionary policy moves in the opposite direction. And they find that it's due to the differences in compositions of incomes and household balance sheets.
In another recent book by a group of IMF economists who looked across 32 economies and found over recent decades that an exogenous 100-point basis point decline in the interest rate lowers the Gini measure of inequality by 1.2% in the short term and by 2.2% in the medium term. So again, evidence across countries that how we think about monetary policy does have this very important effect on inequality and that there's strong evidence for this.
There's also a lot of new research showing how inequality is affecting the transmission of monetary policy to the US economy in several ways. Research by Stanford economist Adrian Auclert shows that differences in the marginal propensities to consume have an important impact on how interest rate changes are affecting aggregate to demand. And in an economy where you're seeing this stark increase in inequality by both incomes and wealth, understanding how the stimulative effects of monetary policy are amplified when it shifts incomes towards individuals who are more likely to spend it is much more important than it was in an era when you had less inequality.
There's also evidence that the transmission of monetary policy is more effective for middle-class households that are more indebted and have adjustable rate interest rates on their debts. So this isn't just about income, but it's also about the question of net assets.
And one other point that I think it would be remiss to not mention is that while traditionally I think that inequality, although it's starting to become more of an issue, more of a concern from researchers and more of a concern of the Fed, this hasn't been as primary of a concern in one place where I think it would be good to see more questioning and especially as we're looking at a lower bound on interest rates, is the effective unconventional policies such as the quantitative easing on inequality through the movement, through asset prices.
So in an era where we're already seeing high wealth inequality, to what extent are these unconventional measures exacerbating that inequality in the short, medium, and long term because of its effects on the asset distribution. That seems to be an imperative question that we need to be thinking about in terms of the costs of running a hot economy. So if our focal point is to think about what monetary policy can do to run a hot economy, but there are costs in terms of exacerbating inequality and this is a major structural problem, then that's something that we need to be spending a lot more time thinking about.
I wanted to end with this slide. Although we're talking about monetary policy today, the alternative to monetary policy to keeping a strong economy, a hot economy of course is fiscal policy. And the biggest, most important boost to stimulative fiscal policy in recent years was the Tax Cuts and Jobs Act, which has enormous and significant implications for distributional questions. So while we've got a little bit of a bang for the buck out of this, it's faded quite quickly.
The fact that this was so unequal in terms of its execution brings to the fore of this question of the extent to which both monetary and fiscal policy to keep this hot economy. Are we thinking about the distributional effects and keeping that front and center? Because if in the long run these policies are exacerbating inequality, then you're only amplifying these larger term structural problems that are keeping wages in check. And I'll end there.
Thank you, Heather. Our next panelist is Doug Holtz-Eakin, who is president of the American Action Forum, and formerly the director of the Congressional Budget Office between 2003 and 2005.
Well, I want to thank the Fed for the invitation today, Mary Daly in particular. It's a great honor to be here. I want to thank everyone in the audience for taking the time to do it, and the panelists who are genuine experts on this for their remarks. And I want to point out just a couple of things. Number one, I don't have any slides. The very first time, I did PowerPoint slides. I did them in front of then President George W. Bush. They contained an error. He noticed it. I got thrown out of the West Wing. I don't slide. So we're going to talk a little bit.
And the other thing I do want to mention in my sort of long and dismal career, I was at one point a commissioner on the congressionally chartered Financial Crisis Inquiry Commission, whose job it was to provide the American people with a coherent explanation of what happened in 2007, 2008. We spent two years and a lot of tax payer money failing to do that. I first want to apologize for that. No one's heard of the Financial Crisis Commission because we didn't do our job. But in the course of that, I did learn and come to appreciate that the Federal Reserve was the single best policy actor in that moment. It did the most to stem the fall and to restore stability in financial markets in the broader economy. This is my first real public chance to say thank you. And for all the people in this room, I thank you for what you did at that time. It was a great crisis. And for those watching the live stream in the West Wing, the Fed remains the premier monetary authority on the planet.
To the topic of the hour. When we were talking about how the panel was going to play out, I got assigned the voice of doom role to talk about the potential costs. Most people did focus on the benefits. I'm not going to disagree a bit with that. I think that's very real. It's worth thinking hard about. It's been just astounding to me, especially through 2018, to every month sit down and look at the jobs report and realize how many people were being drawn into and participating in the labor market where the wage increases were occurring. I'm wrong a lot. To be wrong 12 straight months is frustrating, but we were wrong in the right direction. And I think it is something that I really do have to think hard and appreciate when we think about how we go forward. But it is my job to sort of pick up on some of the things that Doug Elmendorf did. Never be the second Doug to talk.
He's already said everything I'm going to say. And remind ourselves there are some potential trade offs here that we've got to think hard about. And the one he focused most on was the inflation trade off. He crossed out the wrong paragraph. I'll fix that. But it is true that you run the risk of higher inflation. And Jason's presentation highlighted this. A Lot of the recent research has focused on how that trade off may somehow, if not gone away, really, really changed dramatically, and that we don't need to worry about that so much anymore. I've been puzzled by this, as have many people, and at a very simple level you can see it when you do something that got highlighted by one of the questions to Doug Elmendorf. If you just take the PCE and break it into goods and services, from June of 2009 at the end of the downturn and the beginning of the recovery, services inflation has been above two percent, and it has ticked up as the economy has expanded and gotten hotter.
No question about it. It looks like what you would expect to have happen. Goods inflation is the puzzle, and to me, that always suggested that somehow we got to think harder about the global nature of goods markets. They are much more traded, not exclusively, but much more traded and much more influenced by conditions in other economies and the activities of other monetary authorities, and makes the problem much, much harder. There's a really nice set of research by Kristin Forbes at BIS, in their working papers series, that takes sort of traditional Phillips curve inflation models and says, okay, let's think about this from a global perspective. What would we want to do? Let's augment them with exchange rates, oil prices. Instead of pulling them out, put them in and control for them directly, and try to measure the slack in the major economies. And not too shocking, to me at least, that cleans things up, and they start to look like the traditional inflation models did much more so than you might expect.
And that to me is a cautionary note that if the old inflation models remain, the old trade offs remains down there somewhere, and they could pop up. And indeed, that makes it harder for the Fed because they pop up because things go on somewhere else that they don't control and that they may not observe as easily. And that to me is something that's worth thinking a little bit about. So I'm sort of still cautious about the inflation trade off. I'm I think more nervous about the excessive risk taking bubble trade off than some people. This is the 21st century experience. I mean, we had dot com bubble that burst and produced a relatively minor recession, 2000, 2001. That was an equity bubble, and its fall, it was relatively contained. We had a credit slash mortgage bubble collapse in 2007, 2008, produced the greatest recession since The Great Depression.
That was a leverage issue that had dramatic implications for the mainstream economy. And through all of that, the basic research finding was, you can't really identify a bubble. It's hard. It's hard to see it. How would you identify it in advance? And so you really should just wait and clean up afterwards. And the decade of cleanup after the second bubble bursting makes me think that's probably not the right way to think about it. And that the Fed should be more cognizant that low interest rates can produce some excessive risk taking and inflated asset prices, and that we ought to weigh that into the costs of running a hot economy pretty, pretty severely, and keep a close eye on that. I am less convinced, I guess, than Doug Elmendorf that the new financial regulatory regime, which I believe is on average much better and much safer, is cyclically sensitive enough to pick up a run up in a particular asset class and contain it.
And so that I think this risk remains, and is... And it's a place where we really just, we do need a lot more research. And unfortunately, the way to get the best research is to have more data, but I don't want to have more experiences like that, so I'm trying to figure out how to handle that. Let's put that aside. And the last thing I wanted to just touch on, which has also come up already, and which the group thought was worth looking into, is really the spillovers and fall out from a hot economy to other parts of people's standards of living and life, and in particular housing prices and rents, things like that, comes up a lot. I'm not an expert in this, so I went out to look at what literature there might be, and there is shockingly little research on this.
There's an increasingly large research literature that links housing markets to labor markets, in particular the fallout of the housing bubble collapsing, people being underwater, reduced mobility, sort of things where housing markets impinge on the labor market, but less going the other direction. Labor market changes affecting housing markets. But there are a couple pieces of research, one by—I have to use my glasses now and read my notes—by Gan and Zhang in 2013, which looked at Texas data, and basically used metropolitan areas to try and identify labor demand shocks and the resulting outcomes in housing. And you find a nice positive correlation. A gentleman named Saks in Journal of Urban Economics in 2008, similar kind of panel exercise, pretty strong implications between labor demand shocks and housing outcomes. And then more recently, another piece that looked at the heterogeneity across them. I don't think this is a surprising result.
I think this is exactly what I would have expected to happen. It's nice to see people going through and trying to get the magnitudes right, sort of figure out what the implications are. And I guess the major thing I would echo is that when you start looking at the magnitudes, the magnitudes of fallout from the hot economy pale in comparison to the house price implications of land use policies and all the things that the local governments do control. And so that admonition that don't blame the Fed for the high house prices is exactly right. I mean, that's something that other policy makers need to worry about. Second kind of cost that I think we should be cognizant of, and I'm worried about, are risks to the Fed itself. The Congress gave the Fed a dual mandate, which is a really hard thing to satisfy in and of itself.
It is tempting to take the Fed's tools and it's in the conduct of monetary policy, and try to look at implications at different points in the skill distribution or the income distribution or the geographic distribution of workers and Americans. And the evidence today is that temptation is real, because you can see some things that matter, and matter dramatically. But I don't think the Fed has enough tools to satisfy all the distributional objectives in the aggregate objectives simultaneously. And if it becomes the case that the Fed consciously, in its new framework, is going to highlight distributional objectives, it's going to have, I think one of two possibilities. Either A, it will have to have a much more enhanced and vigorous communication strategy to be able to explain the difficulty of doing all these things simultaneously, or to run the risk of being viewed as unsuccessful.
And I don't think that that's a good idea. I think it carries risks as well. And so I think it's also true, as a footnote, that if the Fed is perceived as trying to do those things, it takes these other policy makers off the hook. And I don't think that's appropriate either. I think Doug Elmendorf summarized the efficacy of fiscal policy too gently. They're not as good as he said. And we have to sort of not let that that be viewed as acceptable, because the Fed could somehow go out and fix these problems in some very sophisticated way. So I think this is a great conference. I'm pleased to be able to participate. I do want to emphasize that I focused on the costs. These benefits are real and very, very important. And thank you for the chance to be here.
Thank you, Doug. Our final speaker is Sarah Bohn from the Public Policy Institute of California.
Thank you. Thanks. It's a pleasure to be here and be part of this important conversation. So we're going to end the presentation part of this panel with something of a case study in California, looking at a perhaps particularly hot economy and how the benefits of that have accrued or not to lowest income California. And so I'll be talking about poverty in this state in the context of today's economy. California's economy is booming on many measures with annual GDP growth outpacing that of the U.S. since about 2012 through 2018. California also outpaces the U.S. overall in employment growth for eight years straight, outpacing U.S. rates on that front. At the same time, however, California has persistently high rates of poverty, higher than in the U.S. overall. So about 20 or more percent of Californians are in families below the poverty threshold. And despite some improvements, small improvements over this period, it doesn't appear that there's a massive closing in the gap between California and the U.S. on this measure, unlike we've closed the gap in unemployment rates and other measures of the labor market.
Using the Census Bureau's Supplemental Poverty Measure, we find that California has typically the highest poverty rate in the nation across states, or one two with Washington, DC. So my research focuses on unpacking these trends a bit more to understand economic wellbeing, and especially the role of state policy. And with that in mind, PPIC has partnered with the Stanford Center on Poverty to create what we call the California Poverty Measure, which is a detailed version of the Census Bureau's Supplemental Poverty Measure shown here using mostly Census survey data, as well as some state administrative data, to get more detail on what's happening in California. And I won't get into the weeds of that measure, but I will just assert that I think this is more of a modern way of measuring poverty that takes into account how families share resources, not just their earnings, but also safety net benefits that have changed drastically in scale and scope since the 1960's official poverty measure was created.
We also in our research, or in these poverty measures, account for how families cope with cost of living. Of course housing costs are featuring significantly right now in California, but also considering things like the cost of childcare, the cost of medical expenses. And I think, I hope that this work will provide a few insights for today. I wanted to highlight three that I think are most relevant to this conversation. The first, that the safety net has an important role in smoothing incomes both today and during the recovery, but there's an inherent trade off there for families that are seeing small increases in earnings that might affect their eligibility for programs. I feel like I can't talk about poverty without at least touching on the social safety net. Second, that poverty in California is largely an issue of working poverty, and that is true today.
That was true even before the economy in California was booming as it is now. And third, that how the California housing market adjusts to the booming economy is really critical for understanding whether improvements in employment and wages for low income Californians have really netted out real benefits for them in terms of their wellbeing overall. So I'll take these three points in turn and flush them out a little bit more. So to the first point about the role of the safety net in the California economy, this is a picture of the distribution of family resources across the state. And I should mention that it includes not just earnings, but also benefits from social safety net programs, things like earned income tax credit, as well as cash-like programs, so food assistance, housing subsidies. What you see is that a lot of Californians are in families living pretty close to the poverty line, and we haven't seen this distribution shift drastically since 2011 when we started this research.
In part, that's because the safety net is playing that role in smoothing family incomes or family total resources, perhaps less so than in previous ways that the social safety net was implemented in this country. But nonetheless, today we estimate that 2.7 million additional Californians would be in families with not enough resources to make ends meet were it not for social safety net benefits that are currently part of their family budgets. That's just in a kind of purely accounting, holding all else equal calculation. But a lot of families, as you can see from where the mass of this distribution is, face a trade off where if they do have increases in earnings, as we've seen from previous data shown today, that can yield a decrease in either the size of the benefits that they rely on from safety net programs, or affect their eligibility full stop.
And that is especially true in California because eligibility for many programs, especially federal safety net programs, don't adjust for cost of living, even though wage rates to some extent incorporate the higher cost of living here in California. So indeed what we've seen since 2011 that families near the poverty line but below it are relying less and less on social safety net benefits as a share of the total resources they use to make ends meet, from about 30% of resources in 2011 to 22% today. And that brings me to my second point, which is about the fact that poverty in California is an issue of working poverty. I think a lot of people are surprised to learn just how families make ends meet, even for families at the low end of the spectrum. So among the bottom 10% of California families, two thirds of their total family resources come from their earnings, the other third from social safety net benefits.
That means that employment, both pay, job quality, and job availability, drives poverty to a large extent in California. In fact, what we see is that most poor Californians live in families where at least one adult is working. Those three blue pieces of the pie show those who are in families where at least one adult is working full time, full year, or part time or part year, covering about 70% of those who are poor in California. In particular, so 40% of families in poverty in the state have at least one adult working full time, full year. And this raises natural questions around, is one full time, full year job enough to make ends meet in California? What do those jobs look like? But also, what are the barriers for those who are working only part time or part year?
What would it take to get them to work more hours? Or is it barriers that policy could address, like childcare costs or the need for more skills? So when we looked at the jobs that these working poor Californians hold, hopefully you can read this chart from where you're sitting, these are the top nine occupational categories that we observe today, poor Californians who are working, holding these occupations. And this is just the top nine. It covers about 77% of all working poor Californians. So the first six boxes are about an equal share. So those working in building, grounds maintenance, sales, food preparation, office support, personal care, and transportation and materials moving. So construction, production, and farming occupations are also large, but a little bit smaller than those first six. It's not too surprising when you see this list of occupations among the working poor.
They are also occupations where less formal education is required, which of course is probably the most important determinant of economic wellbeing in today's economy. These are also occupational categories that are growing extremely fast in California. So I've starred those that are among the top five in terms of projected job growth through 2026. And those four occupational groups alone are projected to account for 40% of all new jobs in the state through 2026. So of course those projections assume that things will keep operating as they are, which we know is not a certainty. The fact that these are also occupations where a lot of Californians are working, but in poverty, doesn't mean the growth in these job categories will yield poverty, of course. Families might have multiple earners, individuals might hold multiple jobs. But the wage, the annual earnings information that I've also shown on this chart is not super promising.
So in food preparation, the median annual earnings in 2018 was about $25,000. Same in personal care occupations. A little bit higher, $32,000 in transportation and materials moving. The one exception is in construction, where at $54,000 at the media and in California, there's a little bit more potential there to escape poverty. So while the economy has created a lot of jobs and improved employment outcomes for Californians, they are not, especially for low income Californians, not jobs that necessarily will get them out of poverty. And I just wanted to throw in that we have another strand of research at PPIC looking at career pathways out of low income work, and focusing specifically on what community colleges are providing as one of the key state policy investments to build skills. And based on that research, there is some good news, but I also find it's not super promising that there are strong pathways out of these occupations into higher or middle income level jobs.
So that said, even with the facts of these kind of quality of jobs that have been created for the working poor, we do observe, in the past year especially, strong improvement in family resources, even at the very bottom end of the income spectrum in California. So the bottom 10% of families' resources grew at about 5.8% just over the last year. The growth rate is a little bit smaller as you go up the distribution, but still positive. However, at the same time, the threshold for meeting basic needs has also increased. Across the state on average, the poverty threshold increased about 4.5%. This is driven almost entirely by an increase in the cost of housing. So as you can see, it's increased a bit more in San Francisco. The same is true of other coastal parts of the state, and inland and northern regions saw slightly smaller, but also still increases in what it takes for families to make ends meet.
So looking at these two sets of bars, it's clear to see that how even increases at the low end of the income spectrum might not net out against what it actually takes to make ends meet in California. So what I haven't told you is what this threshold is, and what level of wellbeing we're talking about. And I'll close on that point. So how we measure poverty in this new research is based on really a minimum standard of living, so basic needs and food, clothing, shelter, and utility categories. And what that means in California is, for a family of four on average would need $32,500 to be out of poverty, a little bit higher on the coast, a little bit lower inland and in the northern parts of the state where cost of living is lower.
And as we saw from that distribution I showed you earlier, there are a lot of families near that threshold. We actually estimate that there are as many families, or individuals I should say, within 50% above the threshold as are below the threshold. So in some, that means that there are about 38% of Californians who are either in poverty or near poverty. And the hot economy here in California hasn't changed that fact. It's been pretty intractable that that rate has stayed almost the same since 2011. The pressure on the housing market is one major factor, but also the fact that our booming economy hasn't created new pathways out of low wage work, which we might not expect it to do, but in my mind anyway, is nonetheless the most important and pressing need to really foresee in the future some improvements in poverty in California and ensuring that more of the benefits of a hot economy accrue to lowest income Californians. Thank you.
Thank you so much. So we have time for questions. Maybe I'll kick it off with the following. One thing I'm wondering about is whether the cost and benefit, be it in terms of wage growth or re-entrance in the labor force, depends a bit on whether, on the intensity versus the duration of the hot economy. Whether if having a longer expansion, maybe less intense, maybe because policy rates are a little bit higher, facilitate adaptations by workers and firms and make it more sustainable so that you see more of those benefits.
I'll go first, since everyone else was being quiet. Almost everything I've looked at, it looks just quite like a linear thing. It's a straight line, and each additional year brings more people in. Each additional year improves things sort of like the last year. Now, what's going on underneath that though is employment rates for people with a college education, they went up in the recession, but even at their peak, I think they were still below where people without a college education or high school or less generally are even in a boom. And so you're pulling in, you saw the black unemployment rate, the Hispanic unemployment rate improved more. I didn't show you, you'd see the same thing for high school or less. Mary's a co-author of a paper that tries hard to look for a non-linearity, and may have found it, but I don't know that it was the definitive proof of it.
So the longer it goes on, the better, but is there some particular turning point, I think it's hard to say.
So let me just add to that. Sort of implicit in the question, obviously the longer that you can have these periods of low unemployment last, the better. Which then, to flip that over, that means that you want to do, as Doug pointed out, the last thing you want is to allow some crash in the economy because you've allowed a bubble to develop or this, that, or the other. So it seems, from the data that I've seen, that it is this extension. But the problem is, is that usually once we get to this point, something happens, and we're back at the beginning of the treadmill, and back at the beginning of the line trying to come back down again. So it does seem like this is this very pivotal moment, when we're kind of at the end of our toolbox, to see how we can keep this going, seems incredibly imperative.
Can take a question, or Doug, you want to add to that?
To me, this is sort of just highlights the difficulty of having, sort of thinking about just the Fed doing things. So if you imagined a fiscal policy that had had been assigned the role of producing faster trend growth and really pursued that in a disciplined fashion, you'd have higher productivity, you'd be running higher interest rates. So you'd be less likely to hit the lower bound, less likely to have an inflation shock, less likely to have all the things that forces the Fed to choose between duration and intensity. And that's sort of part of the problem.
Well, Mary. We'll start with Mary.
President Mary C. Daly:
I wanted, because Jason mentioned that he could talk more about this, and I'd like you or anyone else to talk more about this. So one question I get asked a lot when I give talks is, there's a potential cost, it hasn't been mentioned, which is that we pull young workers out of school. We change the incentives to invest in the labor market for experience over the choice to delay work and go to school, and that this ultimately will restrain our potential output growth, but importantly, resign these young workers to a lifetime of low human capital investment. My reading of the literature gives me one view, but I want to fact check that or a robustness check that by asking, I'd love to think about, how do you balance that off, when you net this out?
So I guess I'll start, since you talked to me first. I haven't actually looked at this as much as I should. 25- to 54-year-old men, to me, are the easy group, because there hasn't been large cultural changes, the ones who aren't working don't, they're not the San Francisco person who's in between IPOs or something like that. These are men that seem like they really would want to work and just have given up for some reason or other. Above 55, retirement, trends in retirement, are perfectly legitimate things and that changes over time. And below 25, school is a great investment. One of my colleagues thought the Great Recession was going to result in a period of higher economic growth spread out over decades, because all the people that stayed in school because of it. It was hard for me to be that exciting about the Great Recession as an educational policy.
I guess now to come back, though, to your question, I'm not that worried about people being drawn out of school. I think for a lot of people they also need earnings to be in school. And so for some people it can help them be in school or reduce the amount of time they're distracted from school, because they don't need to work quite as much, because they're being paid more. So I'm not. I guess I'm not. It's hard for me to get that worried about that margin.
I don't know any facts but here's the issue that I see. We've been looking at this sort of various kinds of education and credentialism that people can have, high school, a little bit of college, going to community college. All of these things are valued in the labor market. So more is clearly better. For me then, it's not an issue of just pulling people out of four years at liberal arts colleges to work somewhere else, it's pulling people who would gain a valuable credential or would get some education post-high school and don't. And I think that's a margin that's actually really in play, and that will have big consequences for them. And so if we're going to look at that trade-off, that's the group to focus on, because in the stuff we've done, there are a lot of people out there like that, and it's going to matter.
Your question, Mary, it's not literature that I'm totally familiar with, but I am really familiar with the literature on the other end, about scarring of young people from recessions. I mean, I'm just thinking, I don't see how those both can be consistent at the same time, that during a recession, because people are either going back to school, they're not getting that experience, you see these long term effects on their employment and earnings outcomes. And I don't know how I would gel that with papers that showed a long term decline in productivity, which would presumably be reflected in their employment and earnings for people who graduated into a strong economy. So, those don't seem consistent to me.
So the other data that surrounds this but isn't directly answering it is, if we look at the people who are the real problems in student borrowing, it's individuals who showed up at college and needed some remedial help. That's just an incident flag for future problems. They are far less likely to finish. They're far more likely to have higher debt. And so that looks like a group that in the Great Recession went to school because there wasn't great opportunities. You don't want to lure the wrong people out of the labor market who aren't prepared to do the schoolwork. So I think there's a group there that we need to worry about.
Can I just add to Doug's point about those workers who are 25 to 34, who are in the labor force but in low wage jobs. Governor Brown, former governor of California, deemed them stranded workers, and there's a big push to get them into college or some kind of training post-high school. The problem is that now there's plenty of money to invest in these programs in public universities and colleges, but during the recession, that is one of the first things to cut. So I know we're talking about monetary policy today, but fiscally it's hard, at least in the experience that I've seen in California since the recession, hard to make those investments even if they would make sense during that time of labor market slack.
Other, yes, here.
Mary Pew from Seattle, Washington. Thank you, guys. This was an amazingly insightful panel, and one of the areas that I think about a lot, and you guys really dove into it in a significant way, is just the wealth and income inequality, and how that really impacts our economy. For me, when I walk out of my door at work in downtown Seattle, there's a lot of homeless people. And so when we talk about the economy and the recovery and how everybody is being pulled up, I really think that that misrepresents the reality, and a number of the graphs that were shown highlight that, that a lot of people were being left behind. So my question is, to some degree, we're all talking about this time is different. I'm curious about the imbalances, and what you guys would share, in terms of how you think about the consumption difference that occurs, for example, as a result of the concentration of wealth and income, as we move forward. And even some of the risks, Doug, that you mentioned with regard to this particular recovery, if you can dive into those a little bit more.
Yeah, I'll start, and then Doug can weigh in. I'm glad you asked the question. I'm glad you found that, there's an interesting thread to pull on. One of the things we know, there's a study that we funded last year or the year before by an economist by the name of Uribel. And what he found was that when you look at inflation, which we also think of as an aggregate indicator, maybe we think of it as medical inflation or, we think about it as one price. His research has come to this really interesting conclusion, which is that because of high inequality, and because people are talking about housing and other sorts of goods and services, we know that there's more producers who are really focused on those high end consumers. That's where the money is. That's where you're going to get those big profits.
So there's more competition among producers to reach those high end consumers. And at the same time, there's been less at the bottom. There's been this collapse of the department stores that served lower to middle income families, the JCPenney, the Kmart, all that, and this rise in the dollar stores. But at the same time you've seen this decline in the number of producers who are competing for those dollars at the very bottom, because there aren't very many.
So what what this economist actually found is that that has an effect on the inflation rate, in that inflation is now actually lower for folks at the high end than for folks at the bottom. I'm not even totally sure what you do with that finding. You guys that work at the Fed should think about it. But that's a way that—I mean, I don't know how you would pull that into policy—but that's a way that, as you think about these changes due to high wealth inequality and income inequality that are affecting the margin, that we know that there is these differences in the marginal propensity to consume. We also know that it's driving production in this way, that's having this massive distortion on product markets, on housing markets, on these other aspects of our economy.
So how we take a step back and think about, what are the things that we need to do to address those distortions. And is it focusing on policies that boost education and all the good things at the bottom so that people can build, pull themselves up. That seems really important. But it also seems just as important to think about what high wealth inequality means and to take steps to try to address that more frontally, because it's not clear to me how you're going to solve the problem if you're only focusing on inequality as something at the bottom half. So it's my reaction to your question. I don't know if I got it at all.
So at least in terms of its current expansion, I think the simple version of this is, right now put aside the distribution of household consumption, household spending is doing very well. It's 70% of GDP, and you don't get a recession unless it turns south in some significant way. And so in the scenarios that I was talking about, particularly where you worry about some asset class getting into a bubble, the debt finance ones are the ones most associated with that low end, and that's the one that we see the biggest fallout and impact on the real economy. The experience we had with the dot-com bubble was far more contained. It's really quite striking that if you look at the wealth loss from peak to trough in the dot-com bubble, it's the same as the housing market. That's the same shock to the economy, and the propagation through the debt channels was so enormous, we had a much bigger fall out. That's what I would look for, if I cared about that.
Rosemary, you had a question?
Yes. Good morning. Your presentations have all been great. Thank you very much. I would fall in that range of the newly retired baby boomer and excited about that. My question is related to unions, due to the fact that we've spoken of inequalities now, we've spoken of poverty. What do you see for the future of unions? My experience at United Parcel Service was that in fact we were the union, meaning that we held the highest amount of Teamsters. But do you see that beginning to rear its head, and in fact, what are your thoughts of unions?
I'll start on this one. That was my last slide that I didn't show, was unions across different countries, because I think that's a really important backdrop to this inequality question. It's an important backdrop to whether or not, how the benefits of a hot economy are propagating themselves across the economy, to what extent do workers have the capacity to bargain for higher wages, and then again, how that translates into prices, right? Because if workers don't have unions, how are they bargaining for higher wages? In terms of where it's going next, where unions are going next, there's a very interesting project out of Harvard, out of your shop, that Sharon Block is running with some other folks, called the Clean Slate Initiative, that has been working for the past year to be bringing together all sorts of voices from all across the country to talk about how we think about unions in the 21st century and where they go next.
Maybe you guys in the back know more about this than I do. But I would encourage you to look at that. They're thinking about things like sectoral bargaining, what it would mean to do everything from thinking about new forms of bargaining to what it means to require, to create some rules to require, firms to put workers on boards. So I think that and everything in between. But I think you don't get around this inequality issue without thinking about how you rebalance that voice in the workplace and in our politics.
Bart, you have a question?
Yeah, I wanted to follow up on Heather and Doug's points. Doug raised housing and then Heather raised inflation and inequality. And I wanted to say, if we think about the main monetary transmission mechanism of lowering interest rates in an environment like this, it's largely giving people the opportunity to refinance their mortgage. And refinancing their mortgage is great, because they're fixed payment over the next 30 years. If you just had a mortgage, it's going to be constant and fixed and it's not going to be susceptible to inflation risk. So these are people who have a large part of their consumption not be susceptible to inflation risk, while the other people renting, they're actually seeing an increase in their rent because of the effect of monetary policy on house prices. And there's an enormous discrepancy in the effect of monetary policy across the income distribution, largely based on home ownership rates.
In the street where I live where the income distribution is very mixed, you see this, the poor households who are renters are actually moving out because it's enabled this gentrifying. And the poor homeowners—not poor, these are lower middle income households—they are actually benefiting through an increase in their wealth, and this is enormously impactful on their lives, of course.
I would add to that point. I'm sure other people have other things to say. But the capacity—what I thought you were going to say, but I didn't hear you say—was the capacity to refinance those mortgages is of course also very much tied to the income distribution and the wealth distribution. So many families who should be able to haven't been able to do that, and that's part of what's been a problem over the past few years.
I mean, I'd say that I think that's an important and interesting point. I agree with Doug, it's not like the Fed could be assigned an inequality mandate. It only has one tool and it already has two mandates. One of the two mandates it has, in terms of employment, is far more important for inequality—not far more, is most likely more—important for inequality than any of the other factors that we could worry about. And so I think having these in the background, understanding that maybe it helps motivate how we balance the two goals, how we think about the employment goal, can we find other tools that might affect one and not the other. But I think all of that is second order, compared to the effects of unemployment.
And as Doug said, the fiscal authority should be focused on productivity. I 100% agree with that and would add the fiscal authority needs to be focused on inequality, and the large changes in inequality are because of structural changes like minimum wage or labor unions or changes in trends in education, all sorts of stuff that no monetary policy maker can do anything about.
Let me follow up briefly on this. I'll ask you a question, Heather, because this is interesting. So Jason, you're saying income inequality shouldn't be a mandate of the Fed and that's very reasonable, but Heather, in your presentation you highlighted that the Fed should also pay attention to that in terms of the policies it's using, so whether QE has impact on the wealth distribution that maybe other tools don't have. Can you speak to that a little bit?
Yeah, I mean, and yes. So I take your point, Jason, that the Fed already has two mandates and adding another one, especially a complicated one, seems like a lot, but it does seem as though in order for it to be effective in its mandates, given the changes that we've seen in the economy, taking into account both the effects of the changes in the income and wealth distributions on the capacity for policy makers to act and those outcomes, it seems critically important, because those are the outcomes that we care about, right? We care about full employment, and we care about whether or not it's leading to the kind of hot economy that's going to lead to decreases in poverty. Making that connection seems absolutely imperative.
So the question, and it's so interesting, in your opening remarks, how focused you were on the objectives of the Fed, right? That it's about the unemployment rate and the price level. And you said the word "objectives" like maybe eight times in your opening remarks or something. So it was a bit, how targeted the institution is on that, it's a complicated-
It's the law.
I know, it is the law. Yes, I understand that. So understanding how this change in the economy, in terms of inequality, affects how you think about those objectives, I understand how complicated that is. But these changes in the landscape of what we're looking at, it shouldn't change the objectives, but how it changes how we think about what that, how we look at that, or what tools we're using seems really important. So understanding the outcomes in terms of inequality and understanding what they mean for policy seems like a part of that.
I just want to make sure I understand. So there's one thing, which is to actually target outcomes along the distribution of, say, skills, or income, or something like that. That's not their mandate.
There's another thing, which is the transmission mechanism for the tools they have has changed because of the inequality, and they need to understand that to be effective in pursuing their mandate. That makes perfect sense to me.
And then there's a third thing, which I just want to make sure I understand, which is the tools currently used by the Fed will not be effective with this transmission mechanism, and they need new tools.
That, that third one, I think, is beyond. So I agree with you. The third one I think is a question.
I don't know what I believe, so don't agree with me.
Yeah, yeah, I'm agreeing with your statement, Doug. But I think the third one, to me, seems to be a question.
I don't know that we know the answer to that, but that seems to be where we should be asking the questions that we have in front of us.
Yeah, I completely agree on Doug's second. You need to understand changes in inequality, in so far as they affect the economy. The Fed was deciding between QE and forward guidance, and they both had exactly the same impact on inflation, exactly the same impact on the unemployment. And we had precise estimates that one of them helped inequality more than the other, sure, I'd go with that one. But we can't even quantify what QE does to the average economy, let alone across the income distribution. So that seems worth continuing to do research on. It may be that we'll never know enough, that we'd want to take that seriously in terms of policy.
We have time for one or two questions. Right here.
Back for a minute to the question of education and the idea that, fundamental to, or at least an important piece of, the inequality question relates to education, and just broader, I guess, the future of the workforce as we look into the 21st century. Those of you who are looking at some of the issues, particularly in a hot economy, but I guess in any economy, and the importance of this go on beyond high school to maybe it's a one year certificate, or a two year associate degree, or a four year degree and beyond, are you seeing strategies, either at state levels or otherwise, that are making any difference, in terms of improving that rate of going on, and are there any best practices out there you're seeing that address this in maybe a more effective way?
Can I not answer your question but complain for 10 seconds? So we do something called the National Assessment of Educational Progress, NAEP scores. We look at fourth and eighth graders. We test them on reading at grade level and math. And if you look at those scores, a quarter to a third of fourth and eighth graders are seriously deficient in math and reading. They cannot read even close to grade level. They can't do math. A quarter to a third of our young people. And these numbers aren't declining. We're staying at a quarter and third, and no one seems to care. We are baking in the cake a serious inequality problem, and what I view as just an immoral situation for those kids. If you had to pick one education problem to fix, fix that one. I don't know how, I'm not an education person, but that seems important to me.
I'll just add, there are some historic, I think, reforms going on in California in higher education policy, to the remediation point that Doug made earlier. Mandates now across the state that students not be placed in remediation and community college, or very short, for very short periods. So it used to be that 80% of new entrants to community colleges entered into remediation in math or English, and very, very few of them ever completed any credential. And now, they are going straight into college level coursework with supports, and we're just starting to see the effectiveness of that. It's a little bit TBD, on how their long term outcomes will look, but I think that's one example, on the more less dismal side of what's being done to try to improve educational outcomes across the board.
Let's take one last question. Nicolas, you have one question?
Nicolas Petrosky-Nadeau of the Federal Reserve Bank of San Francisco here. I was curious to hear your perspective on the demand side of the labor market. We haven't heard much about the costs of benefits, the hot economy, on the difficulty to hire, but maybe what are your thoughts on how it's affecting firms' incentives to train workers, or the alternative, to substitute workers with capital through automation, for instance.
When I was in government, I think I first started getting parades of people coming in, telling me about how horrible labor shortages were, in, like, 2010. I'd always ask them, where are the wage increases you would associate with these terrible labor shortages? Now we're seeing some of those wage increases, but it's been a while.
And this may be a little bit similar to the way I'd think about it in the answer to Mary's question. If you had a temporarily, unsustainably hot economy, and that brought people out of school and then it went away and they were sort of stuck, or that led to a huge substitution of capital for labor and then they got stuck, that would be a bad thing. If it's sustainable, I don't think you're going to see a big capital labor substitution based on one or two years of data. We need capital deepening. That increases productivity. Productivity, all else equal, is associated with higher wages. We are swimming upstream against the slower productivity growth, as I said before. So I guess I'd view the way that the demand side has to adapt to a hotter economy as all good and exciting things for the economy, not something to be worried about or a downside.
I don't have any empirical data, right? The anecdotal things that I would share with you would be, there isn't one topic that's more common when people come through the door at the American Action Forum than workforce training, skills upgrading, re-skilling, however they want to label it. All the trade groups, lots of companies, the National Association of Manufacturers just took over this program that Toyota had for a long time to do in-house training. When people show up, the first thing we want to do is have the government pay for the training of their workers, but they're recognizing they have to train the workers. And if this stays hot, that's going to be an imperative.
Well, this has been just a terrific conversation. Thanks. Thank you so much for agreeing to participate.