Bloomberg Audio Studios, podcasts, radio, news.
Anna Wong has been definitive at Bloomberg Intelligence in Market Economics. She is in conversation with Kevin Hassett of Greenfield, Massachusetts, and of the University of Pennsylvania, now at the White House. Let's listen.
100% of its sales in the US. We're the only customer for them and they have like 1% of the US market and there are lots of other firms selling the same product into the US. Well then if we put a tariff on that guy then it's really going to cause some harm to that firm. And so what we did is we actually went through all the different tariff classes or, yeah, I guess what are the codes called? HS. Yeah. Okay. We went through the ball and then figured out the relative elasticities and then maximized the sort of or minimized the potential harm. On this one, one of the things that we've learned is that one of the reasons why the US has a really persistent trade deficit is that we have a few trading partners that have basically a policy in their country to create jobs and dump product into the US in order to produce political stability at home for them. And so our intuition early on as we're thinking about what's the next step given the president's intuition about how tariffs can be an important part of optimal revenue policy is that if we put tariffs on say China, then their supply is really inelastic because by definition it is the most inelastic thing in the world because they've been just throwing stuff at us year after year after year. Then they would have to cut their prices to continue to sell enough stuff so they could create enough jobs. They had the political stability that they wanted. And I think if you look at the record, that's more or less been the case since the tariffs came on that for the countries that have big persistent trade deficits with us, they've ended up having to cut their import prices by so much that there's very little visible effect on inflation here in the US. So when I look at the schedule of tariffs right now, China is at 31%. And Brazil and India are at 35%. Which is higher than China.
And is this a transition phase? Is there where the end point of this trade policy in this administration say by 2028 the tariff schedule and the relative tariffs around the country will be different than what we have served now and what's the endgame?
President Trump, you know, we'll decide what the endgame is. He'll look at the deals that are being presented to us and decide if the deal was good enough. You saw for example that we just had a big reduction in the tariff for Switzerland after a very acrimonious phone call between the president and the Swiss that led to the very high tariff rate. And so I think that there's a negotiation underway. I think for many countries the negotiation has finished and for some countries there's still work to be done.
Do you think after all the negotiations, the trade deals and adjustment in a tariff schedule, we might be converging to your algorithm in the first Trump administration?
No, I think that what we'll be converging to is something like a border adjusted business cash flow tax. Because in the big beautiful bill, one of the things we did is we expanded the various export subsidies for multinationals. And so if you have an export subsidy plus a tariff plus a payroll tax, then you basically have a border adjusted business cash flow tax. And so I think that in equilibrium, what we're seeing is that we're getting lots of revenue. It's a few hundred billion already this year. And it's not causing a lot of harm. GDP now is at 4%. Inflation has been trending slightly down. And so I think that it's working the way it would if it were a border adjusted business cash flow tax.
And what have you learned about optimal how tariff fits in this optimal tariff as well as optimal fiscal policy?
I think that there's a race to produce in the US to avoid tariffs that is very beneficial for US workers. And if you look at capital spending year-over-year, we're having one of the best years ever. And that's even before we go through the trillions of promises that we've made in the trade deals. And so there's that which is the capital spending effect. That was somewhat of a surprise to me how big it was. Because I again thought that what you might see is price adjustments and so on by the inelastic suppliers and then the consumers wouldn't really see much of a price increase. But I think one of the things we're seeing is a massive onshoring of activity unlike anything I've ever seen. So I think that it's really been a very net positive policy despite what conventional economics has been teaching us since we went back to the textbooks when I was in grad school.
Okay. So I think on Wall Street, most of the firms, many of the firms have come up with this cost burden of tariff composition of around 5%. Exporters cut discounts 70%, 60 to 70% absorbed by domestic firms and the rest, 20 to 30%, with the tariff showing up in consumer prices. Do you disagree with that sort of breakdown?
Well, I think it will depend on a product-by-product kind of analysis and I think that it's still work in progress to figure out what the final story is. But again, we've got high growth and no pickup in inflation. I think that the Wall Street folks or a lot of Wall Street folks who I respect and read in the spring were saying that we're going to have stagflation, that we're going to have four or 5% inflation because of the tariffs and no growth. And so I think that allocating the current state of the economy and the impact of tariffs on it from people who told us that we'd have a recession with high inflation with those models is probably not the best thing to do.
Yes, certainly Wall Street have been caught wrong on their forecast of an inflation surge which did not happen. In fact, I remember the consensus upon Liberation Day, everybody revised up the core PCE inflation to well over 3%, even some closer to 4%. And now we are on track to hit 2.9 or 3% at the end of this year. So for sure the consensus has been wrong on that. So recently there is a viral San Francisco Fed paper that looked at a hundred years of history of tariffs and they found the net effect of tariffs tend to be deflationary. However, they also found that because it's deflationary because it's an aggregate demand shock more so than a supply shock, and hence ultimately it raised the unemployment rate. And so far this year, when we began the year, the consensus on Wall Street was 4.1% unemployment rate. Currently it looks as if we are heading to 4.3 to 4.4 at the end of this year. How much do you think of this slower hiring and increase in unemployment rate is attributable to trade policy or what other policies?
Well, I think it's actually quite a mistake to look at a 100 years of history to learn about today's situation. To do that you're making an underlying assumption of what econometricians call ergodicity. The idea that the data generating process is the same over time so that you can identify it in this period and then use your knowledge to think about the next period. I think that a hundred years ago, if you wanted to move production from one place to another it was almost impossible. You'd have to get it on the Queen Mary or something and move it around. And now it's really, really easy. A hundred years ago we didn't have a telephone, or maybe we just started to. But now we've got the internet and we've got AI helping us decide how to move stuff around. We've got a capital spending boom that would have been impossible a hundred years ago. And so I don't think that history in this case is going to be that useful because it's clearly a point where the assumption of ergodicity is not going to apply. And so we're seeing massive movements. You've seen the big drug companies, movements of production to the US that will benefit US consumers and especially US workers. We're seeing real wages already this year going up, recovering maybe about a third of the ground that we lost under Joe Biden with the high inflation. And I think that those real wages are going up because we're getting a lot of onshoring. It's not just because of the big beautiful bill's tax cuts, but also because of the onshoring from the tariffs.
Well, you know that conclusion is not just from that paper based on 100 years of history. Actually back in 2018, when we look at the historical FOMC presentations, there was a staff presentation that used the internal workhorse model of the Fed, the Sigma model, to evaluate your tariff algorithm back in 2018. Suppose there's 15% tariff on intermediate goods only and 15% on consumption goods only. And that model found that if you impose tariffs on intermediate goods, it is also over time deflationary and it boosts the unemployment rate by a little bit. In which case, how would you explain then the unemployment rate increasing from earlier this year 4.1 to 4.3 or 4.4% later this year? What are the drivers?
Yeah, I mean it's such a small movement that it's hard to say exactly, but I would say that one of the interesting puzzles is that we're having a productivity boom that is really unprecedented. When I worked with Alan Greenspan back in the '90s, it was just at the dawn of the computer age really. They didn't even have Netscape Navigator when I first started working at the Fed. You had to use something called Gopher. It really wasn't that good. Greenspan had an intuition that the computer was going to revolutionize the economy and give us lots of growth, it was a big positive supply shock. And so when the unemployment rate actually kind of went down, we had no evidence of the productivity. Remember Robert Gordon was saying it's everywhere but in the data. But Greenspan decided not to hike rates because he believed that we were in the midst of a positive supply shock that was perhaps unprecedented. And we had a kind of five-year run, one of the best five-year runs we've ever seen, even all the way to the point where the US fiscal situation was in surplus. And so I think that it's one of the great moves of Federal Reserve history that Greenspan saw that it was a positive supply shock that would be both pro-growth and deflationary.
Yes, I have more stuff about that down. Okay. So I say that to reference today. It's really interesting that there's evidence in peer-reviewed papers or not quite peer-reviewed yet but at NBER, which is kind of a lot of my friends at NBER by the way, it's a secret of NBER authors like myself is that you'll very often take a paper and submit it to a journal and get the referee reports before you put it into the NBER working paper series because if you got some stupid mistake the referees will catch it and you don't want to embarrass yourself in front of everybody by just throwing it out there. And so actually the NBER is kind of almost peer-reviewed, but there's a lot of stuff showing big AI gains. And the thing is that we also see it in productivity. And so if you think about the difference between say 1996 and today is that we've got this revolutionary new thing that probably is going to be a big positive for productivity. And you know, Steve Neri and Dan Sichel, your former colleagues at the Fed, had a famous paper that you only see productivity booms after the benchmark revision, right? So you never actually experience it while it's happening in the data. And this AI boom is significant enough that we're already seeing it in the data. Probably the benchmark revision is going to revise it up. And against that backdrop, I think that the open question is how do we have high growth and then less growth than employment than we expected. And I think that the tariff could be part of the story but I think that probably is a very small part of the story.
Yeah. I mean when we looked at the earning transcripts from this third quarter, the earnings have been surprisingly positive. In fact it was like the best quarter ever surprises. We use AI to extract some of the themes and we did find that one of the dominant themes is that there are other stuff happening outside the trade policy that's helping offset a lot of this tariff bill. But I don't mean to cut you off, but I just want to highlight how important what you just said is. The way to think about it is that trade is about 15% of the economy and the rest of it is like 85% of the economy. And so you could do a lot of stuff in trade that has effects on the trade space that is completely overwhelmed by the 85%. And 85% of the economy is getting this massive productivity change and it'd be hard for a large change to 85% to not overwhelm something you did to the 15%.
Yet at the same time, 66% of the economy is labor share and a lot of academic studies are also looking into whether AI is already causing low hiring rate among young graduates. Whoever is graduating from college this year is probably in for a bad deal. So I think there is a puzzle out there which is if you have GDP growing at 3 or 4%, at the same time you have a lot of jobs, a sense of job insecurity in this labor market, and the labor market is lagging these other GDP growth indicators. The question is can this GDP growth be sustained at this three or four percent while the labor market is going sideways basically with labor share probably going down because of productivity gains in the future.
It could be that. And also don't forget that there's been a big reduction in illegal immigration and employment from that aspect. And so the population growth is pretty low compared to what it was when we had tens of millions of people crossing the border. And so the breakeven jobs number for the payroll survey is down quite a bit to maybe around 40 or 50,000.
Yeah. Between 30 to 60 down.
Well, I was 40 to 50, you were 30 to 60. I think we're about the same range, right? And so that's another thing that we could have big increases in productivity and big increases in output and a healthy labor market with job numbers like the one that you forecasted and we'll see whether you're right or not.
So 54 sounds good to you.
I'm saying that it's not 100. It's that that's enough to sort of keep up with population. You'd want stronger growth than that for unemployment to go down, but the unemployment rate is a little higher than what's probably the equilibrium unemployment rate in a full employment economy. But not a whole lot.
So, Kevin, going back to this productivity boom. I recently tweeted a couple pictures that seem to suggest some emergent micro evidence of how AI is boosting business formations. And St. Louis Fed also has a piece out showing how AI adoption is correlated to a rise in sector specific unemployment. So there's evidence that AI is working through at the micro level. But then the response I got made me realize that it is actually a really polarizing topic with half of people probably saying AI is a bubble. Half of them would say that there's no evidence that there's a productivity boom yet, that productivity is still on trend and the improvement reflects more of a cyclical thing that happened post pandemic as opposed to this sustained AI-driven productivity. So if we are going to touch on you being one of the leading candidates for Fed chair and going back to 1995, 1996, those two critical years at the Fed, you mentioned that Alan Greenspan was really the only one of the few who saw evidence of productivity boom. I reread the historical transcripts in December 1995 that particular FOMC meeting where half or more of the FOMC were agitating for holding rates constant and in 1996 many of them wanted rate hikes. And somehow Greenspan was able to convince all these skeptics that something is afoot without the data to support.
Let me talk about this because there's an interesting anecdote that I don't think is ever made public about this. Because computers were improving so quickly, it was decided by the Bureau of Economic Analysis, a lot of great professionals over there, that they needed to what we call hedonically adjust the computer price. Which meant that if you had a computer chip that was twice as fast as the one from last year and it costs the same, then that's like a big reduction in the price of computing. And so in a typical year right around when we were starting to have to make the judgment on this, the computer price was dropping at about 16 to 20% at an annual rate and there was so much computer investment that real GDP was getting a real big kick from the fact that the deflator for computers was dropping so much.
Yes. So I was in charge of that part of the world for the staff because I spoke a lot to Greenspan and he said to me once, how do they deflate communications equipment because communications equipment is basically like computers now. And I went back and looked and communications equipment rather than dropping 15 or 20% a year was going up like 2 or 3% a year. And so I called over because they're actually very open about the technical people at BEA, right? And I talked to them about why is that? And they said that when we're making our deflators we use something called a matched model approach. So if something we sold it yesterday and we're selling it today, then we look at the price change for that and that's where we get our deflator. And we don't change the bundle of things that we use the matched model approach for very often. And it turns out that there was a lot of communications equipment produced in the US for developing countries that were kind of really old-fashioned analog things where you got like Josephine moving the wires and stuff to connect the phone calls. And since we had this really old equipment that we were exporting to developing countries, they were using the price of that to deflate communications equipment. It sounds like a very tactical matter. But then Chairman Greenspan said, 'Well, Kevin, what if we deflate communications equipment with a computer deflator and instead how much more would GDP be going up?' And it was like about one and a half percent because of that. And so he was kind of like, 'Yeah, you know, we're really growing way more than we think. We've got a serious measurement problem.' And then if you extrapolate from that, if you were hedonically adjusting everything that gets like a computer to make it better, then you're putting a deflationary force into the economy that's not being measured. And so I think that the communications story was a key moment where everybody at the board had an aha, yeah I can understand why Alan is so convinced about this.
So I was going to ask you, it's interesting you were in the price and wages section.
No, I was the business fixed investment person.
Oh, I see. And what other sections were you or have you?
Wow. That was a very well placed for that.
Well, that was my area, right? Capital spending and taxes and investment.
He did mention in the transcripts the rapid deceleration in the prices of high-tech prices.
But I mean Greenspan is the pioneer of alternative data. He talks about men's underwear sales for recession, the weight of GDP, remember that? Lipsticks. You had people had to calculate the weight of GDP, I'm not kidding. So he's a chairman who's really in the weeds in the data and very creative as well. But let's not underestimate the challenges he faced on the committee where half of the committee was against this. How did you have any insight then on how he was able to win over the Fed staff, which is very influential? Back then you have the green book and the blue book. Now I used to write the green book twice a year.
Yes. And do you know what color is the book now?
Is it the teal book? Why did they do that?
To mix the green and the blue together that gives you teal.
