T. Rowe Price经济学家：美联储偏重通胀，AI有助降通胀

I feel like if we had this conversation a few weeks ago, it would be a little bit different about where we thought rates are going right now, at least if you look at the Bloomberg terminal, about a 50% chance right now for a rate hike in September is what traders are betting on. And now you have a lot of economists saying that even if we get that hike in September, it might not be the only one. Why? Well, I think this is a very interesting question.

And some of that is to do with the data and the volatility in the inflation prints. The fact that we haven't had that much progress on inflation over the last five years, but a lot of it has also to do with some misunderstanding or lack of clarity in terms of the fed reaction function under the new fed chair, Kevin Warsh. And so what has happened since the weekend is that now we have a bit more visibility. We don't have forward guidance, and I think it's clear that he's not going to go down that path.

But we have some guidance with regards to his reaction function. And I think the market is reading a lot into his statement that financial conditions are actually not that tight, that 2% inflation is still the goal, and we haven't seen sufficient progress in bringing inflation down, even with the last two reports that have trended lower and and showed some progress. But he indicated on Friday that that's not enough yet.

When you look at the market reaction to what he said, do you think the market got the right message? I feel that they did. Uh, the thing is, was did, uh, his reaction function, he did not communicate that very well between the June and the July Fed meeting. I think he made a mistake to say that financial conditions have tightened, and therefore the market is doing the work for us. And so the market needed to see this, uh, reassessment and this recommitment to price stability and to the 2% inflation target.

And now I feel they have the right, uh, signal now, what's important between now and the September FOMC meeting, we have two data reports. We have CPI and we have payroll data. The decision in September remains data dependent. But the hurdle I think has become Quite different than it was before Friday. What's the balance between those two? That labor market report coming this Friday a CPI report. Uh, in the in the days ahead after that.

I mean, there are a lot of people that came out of that speech at Jackson Hole and said that this isn't a fed. That sounds like it's, you know, following a dual mandate. It looks like, at least for the short term, there is one mandate, and that is bring inflation back down to that 2% target. Is that how you read it? I think that's a great question and I do agree with you. So what we heard on Friday is let's look at what's happened with wages and the labor market.

While the labor market is in full employment and wage growth is decelerating. But force very clearly said I'm not taking signal from that. Wages are not a good predictor of forward inflation. What I'm looking at is the breadth of inflation. We have 54% of the PC basket. That has been about 3% for quite some time. That's the kind of bread that he seems to be uncomfortable with. So I think the reaction function is biased towards inflation right now.

And the fed is acting like a single mandate. Central bank, at least temporarily. As an economist, though. Do you believe that with regards to the idea that wages aren't necessarily the best gauge to follow? I wouldn't agree with that assessment. I would say that wages are not a risk to the inflation outlook right now, because productivity growth has been quite resilient. And so that explains part of the wage growth that we've seen in recent quarters.

And as long as the economy is becoming more productive, even higher rates of wage inflation are not necessarily inflationary. So I don't agree with your statement that wages don't matter for the outlook for inflation. But I do think that wages are not an upside risk to inflation going forward. Does the US unemployment rate. Does that matter as much. Is that a proper signal or is it maybe a missed signal? The labor market in the US has been buffeted by many structural forces in recent years, and I think those ongoing structural forces are driving the unemployment rate.

That's why it's been very important to look at the labour market in its entirety, and to look at hiring rates and vacancy rates, as well as the unemployment rate. Because when I look at the unemployment rate now, it is true that is lower. But a lot of that is being driven by labour supply and demographics, ageing population, slower migrant flows. It's not necessarily a sign of a tight labour market. So then you go to wages and wages are moderate.

Then you go to the vacancy rate which is low going to pre-pandemic levels. You look at hiring rates. They're not very elevated. So we don't have a hard labour market right now. You mentioned productivity. And I am curious how productivity gains, uh, sort of address the inflation situation. And I assume those productivity gains are coming partly from some of this AI adoption, if you will, and implementation. Yes. And I think we need to look at productivity over a long period of time in the near term will have the, uh, easy productivity gains.

The way we measure it is output per worker, GDP per worker. And so when you're not hiring a lot in the economy, but output or GDP is increasing because you are investing a lot in CapEx growth is high. Then you'll have a mechanical improvement in productivity growth. But then over the medium to long run, what will matter is total factory productivity and the interaction of this new technology, AI with labor. And I think those gains are still ahead of us.

So I'm actually quite optimistic about the outlook for U.S. productivity. Stronger productivity growth also means we can look forward to higher GDP growth, both real and nominal, without it necessarily being inflationary. So when I'm looking at the outlook for next year and beyond, I'm quite positive that we'll be able to bring inflation down to 2% even without interest rates being too restrictive. Because of this productivity gains, with regard to how AI feeds into that, I mean, some people are saying the AI itself is inflationary.

It may be disinflationary at some point. You don't buy into that. I don't I think this kind of debate doesn't fully take account for all of the effects of AI in the economy. So this debate that, uh, in the short run I will be inflationary is looking at specific CPI components like software, like computers. And it's true those are increasing, but they're a small part of the basket. However, this debate is ignoring completely the effect of AI on the labor market.

The fact that we have this labor enhancing technology, the fact that workers maybe don't feel so confident to switch jobs right now, or to us for higher wage increases, I think that has an effect on inflation on the other side. Now, it's much harder to quantify it, but if I had to bet on one outcome, I would say that the disinflationary effect of AI will be much larger, both in the near and medium term. With regards to some of the structural issues in the economy, there's been a lot of focus on the U.S. deficit, uh, its debt load, uh, and more importantly, how the Treasury is managing that.

Uh, does that sort of work in conflict with maybe what we think Kevin Warsh wants to do with regards to both his communication and his desire to get inflation back down to that 2% target? So I would say that given the reaction of the long end of the US Treasury curve over the last 2 or 3 months. The three are very, very closely related. And what we notice every time Warsh comes out with a message that is tough on inflation.

That emphasizes Fred Fed credibility. We see the long and yields coming down, term premia suppressing. So I think those three things are very closely together. And trying to achieve lower yields by cutting in

原文超出正文长度上限，此处截断——上游还有内容，完整版见上方「原文 ↗」。