What an Expanded Beveridge Curve Could Tell Us about Soft Landings

Tim Lloyd: Welcome to Timely Topics from the St. Louis Fed. I’m your host, Tim Lloyd. And with me today is Paulina Restrepo-Echavarría, an economic policy advisor at the Federal Reserve Bank of Saint Louis.

She’s been kind enough to give some of her time to help us better understand a topic that’s been on a lot of folks minds right now: What is a so-called soft landing for the economy, and in particular, how something called the Beveridge curve relates to a soft landing.

Paulina, thanks for joining us.

Paulina Restrepo-Echavarría: Thank you for having me.

Lloyd: Before we get into some of the economic factors that go into determining what a soft landing is, can we just take a step back and explain to folks, what is a soft landing?

Restrepo-Echavarría: A soft landing is basically when, the central bank manages to lower inflation without generating unemployment.

Lloyd: And one of the factors we hear a lot about is something called the Beveridge curve. And I should just say for folks has nothing to do with drinks of any kind. It’s spelled, B-E-V-E-R-I-D-G-E, the Beveridge curve. What exactly is that and why do we hear about it so much when it comes to the idea of a soft landing?

Restrepo-Echavarría: That’s a very good question. So, the Beveridge curve is the relationship between the vacancy rate and the unemployment rate. So, it basically tells you what happens to unemployment when vacancies go up or down. In a nutshell, that’s what it is.

But why is it so important for understanding a soft landing? That’s the key question.

Take for a moment the mechanism through which monetary policy operates. If the central bank, or in this case the Federal Reserve, were to increase rates, then that’s going to make access to credit more expensive. And imagine that, for example, a firm needs access to credit in order to pay its wage bill. Once that’s going to be more expensive, more likely they’re going to reduce the number of vacancies. This means that, as monetary policy transmits through the economy, what it will generate is a tightening in credit that can end up reducing the number of job openings or vacancies that firms have.

So, the Beveridge curve will, in principle, tells us by that reduction in vacancies or job openings that we’re seeing how much unemployment that’s going to generate. That’s why the Beveridge curve is really important to think about a soft landing, because that would let us know if an increase in interest rates that we need to impose in the economy to lower inflation would generate a lot of unemployment or not so much unemployment.

Lloyd: Okay. So that’s what the Beveridge curve is. But you and Dallas Fed Macroeconomist Anton Cheremukhin recently expanded on what you were just describing to include two factors. Tell us about that.

Restrepo-Echavarría: Yes. So, something that we think has been very important for this period of time, the last two years. We’ve observed a very tight labor market. What does that mean? We’ve observed very high levels of vacancies relative to unemployment. And the Beveridge curve has behaved in ways that are very unusual.

Why is that?

Because right after the pandemic, we have observed that the Beveridge curve has been flat, meaning that, you have to move vacancies very little to change unemployment by a lot. This was at the very beginning of the pandemic when there were a lot of layoffs. But then we’ve seen that the curve has had a couple of vertical segments, which means that you can lower or increase vacancies by quite a bit without generating any change in unemployment.

And this is very, very puzzling because usually, or historically, the Beveridge curve has a negative slope. Meaning that you’re just going to see a negative relationship between vacancies and unemployment, meaning you move vacancies somewhat and unemployment will move by that much or so. Usually, you know, you may see slopes that are kind of equal to one, meaning that there’s a 1 to 1 negative relationship. But in the recent period, we haven’t seen that. It’s looked very, very unusual.

So, we—my coauthor from the Dallas Fed, we started thinking, well, what can explain this very puzzling behavior of the Beveridge curve? And we realized there’s something to be acknowledged here and that is that not all vacancies are made equal. When a firm decides to hire a worker, where do they hire from?

They don’t always hire from unemployment. Sometimes they hire unemployed worker. So, in other words, they might open a vacancy for poaching someone from another firm, or they might open a vacancy to hire from unemployment.

So, once you think about that, and imagine that the total number of vacancies that we observe from JOLTS, which stands for Job Openings and Labor Turnover Survey which is a survey collected by the U.S. Bureau of Labor Statistics, tells us when we see the total number of vacancies in there, those vacancies include both poaching vacancies and unemployment vacancies. And so, when we think about the Beveridge curve, if you kind of think about it, poaching vacancies have nothing to do with the Beveridge curve because poaching vacancies don’t affect unemployment because those affect only job to job transitions.

So, if poaching vacancies go down, then your opportunity to move from one job to another, from one firm to another goes down. But you’re not going to be unemployed because you’re already employed somewhere else. So, really, the vacancies that should be taken into account in the Beveridge curve should be those that are open to higher unemployed workers. So, in my work with Anton, we come up with a methodology to be able to estimate and measure the fraction of vacancies that are open for poaching versus the number of vacancies that are open for hiring from unemployment.  And we find actually, that since 2015, the fraction of poaching vacancies has been going up drastically to the point that they make up 80% of total vacancies at some point in time.

That’s a huge increase because it used to be that prior to 2015, the fraction was 50/50. So, more or less half of the vacancies were for unemployed, half of the vacancies were for poaching. But that number has gone up to 80%.

That has a big implication for monetary policy because imagine that through the channel that I just explained to you, in which, you know, firms have, let’s say, restricted access to credit or credit is more expensive, and they reduce the number of vacancies that they open. Well, if we think that monetary policy is going to affect then a symmetric way, both poaching and unemployment vacancies, then most of the vacancies that are going to be sacrificed, quote-unquote, are going to be poaching vacancies. That means that unemployment is not really going to be affected.

Lloyd: So, if I understand you correctly, what you did was  go back and revise the model, if you will, to help us better understand this historically at least unusual behavior that started after 2015. But again, help us connect this back to what a soft landing might be.

Restrepo-Echavarría: Yes, so that’s a very good point. So, for example, let me be more specific about that.

Take for example March of 2022. Job openings were at 12.18 million. Unemployment was a 3.6%, CPI. Inflation was at 8.55%. and the Fed funds rate was at 0.2%. Now, let’s move forward two years into March 2024. In March 2024, job openings were at 8.48 million, unemployment at 3.8%, CPI inflation at 3.47%, and the Fed funds rate had gone up to 5.33%.

So, notice that here the Fed funds rate went up 513 basis point, inflation came down 508 basis points. Job openings went down 30% but unemployment up only 20 basis points by March 2024. So, if you think about it, the increase in unemployment of only 20 basis points shows us that the relationship between, vacancies and unemployment is not that negative. Meaning that, yes, 30% of job openings were cut down, but this only generated very little unemployment. How is this even possible?

Here’s where our theory comes in. Because the job openings that were quote-unquote sacrificed were poaching vacancies, were job openings for poaching another worker from another firm. So, then that allowed the increase in unemployment to be much lower. And that’s exactly the definition of a soft landing. So, what we’ve seen between March 2022, on March 20th, 24 is that the definition that we have as economies of a soft landing, because of the numbers that I just gave you.

Now, the question really is, is this going to continue or not? And that’s the big question.

Lloyd: Right. And that’s where the Beveridge curve again comes into that consideration.

Restrepo-Echavarría: Yes, exactly. Because whether the slope in the Beveridge curve changed forever or not would tell us if this can continue going forward or not really.

Lloyd: Right. And I do want to pick back up on one thing that you talked about earlier and that is that we’re looking at sort of numbers in and around the pandemic, if you will. But let’s go back to 2015, why do you think we started to see that increase in poaching of already employed workers?

Restrepo-Echavarría: Yes, exactly. And that’s the question that is also going to allow us, to figure out, I guess, or say something about, you know, what’s going to happen from now on. Will this trend that we’ve seen continue happen or not?

What we show in our paper is that what has changed since 2015 and over time to generate this increase in poaching vacancies is the profit to cost ratio. What do I mean by that?

So, there’s something that has made poaching vacancies to be more profitable, or better for a firm to open, and so the question is what can be generating this. So, this increase in the profitability of these types of vacancies can be due to the expansion of online search tools as well as AI, for example.

Something else could certainly be the expansion of available temporary or remote work.

The rising market, concentration and markups, which has been documented in the literature by several authors at this point, as well. Or monopsony power of firms. Any of these different things have been determining this increase in poaching vacancies. Going forward we should observe poaching vacancies a big majority of, you know, of total vacancies, a big fraction of them. However, if really the explanation is not along those lines, but let’s say that [it] is about a sustained reduction in mismeasurement of vacancies, which means that before 2015, if firms weren’t really advertising their vacancies, and all of a sudden they started reporting their job openings, then that’s a totally different ballgame.

But for now, to get back to your question, what we think is going on since 2015 is that this increase in the profit cost ratio, which is a mathematical object that we can show is actually the reason why these trends started going up. If that is maintained, then, you know, this relationship between unemployment and vacancies is going to keep being weak. If it doesn’t, then it’s a separate thing.

Lloyd: We’ve been talking a lot about the beverage Curve as it relates to, a soft landing and in particular, your work to explain some of these historically unusual trends that we’ve we’ve seen since 2015.

But, I do want to ask, are there other factors that economists often take into consideration when thinking about the soft landing?

Restrepo-Echavarría: Yes, that’s a good question. The indicators of a soft landing are employment and GDP. So, if for example, you see interest rates going up or they are high, and you see no meaningful increase in unemployment and a decrease in GDP, that would be a soft landing.

Now if you this is actually what seems a very a fairly simple question is quite complicated because different economists will have a different perspective, even though the end product is the same, the data points that you would look at the end are the same, you would get people that would argue that you can look at financial conditions also because, you know, financial conditions can transmit through the rest of the economy. If  financial conditions seem to be very tight, then that can be an indicator that you’re not going to be able to pull off a soft landing. And that’s if you want to assess in the process if you’re going to be able to pull off a soft landing.

But at the end of the day, the end product is unemployment and output.

Lloyd: You know, one of the great things about research is that as you kind of dig into these historically unusual trends and begin to create new ways of looking at the Beveridge curve, like you had mentioned before, there’s lots of other questions that arise ,when you start answering one question and another one pops up. So, what’s next? What are you going to look at next, as it relates to the Beveridge curve?

Restrepo-Echavarría: So, for now, we’ve been focusing specifically on labor markets and how when you think about the process through which employees matched with firms happens. And we’ve only done it in this I mentioned. Now, I’ve talked to you about the relationship between the Beveridge curve and the soft landing through the mechanism that we observe monetary policy transmitting into the real economy. What do I mean by that? We started by discussing how when the Fed funds rate goes up, then this will make credit more expensive for firms. And this will make them reduce the number of vacancies. That part is not in our model yet.

So, we know that the mechanism is there. That’s exactly how monetary policy is supposed to work. But we don’t have that yet in our paper. So, we’re working on, let’s call it a full-blown model where we can incorporate monetary policy and how that actually transmits to firms. And then firms will make the decision of which type of vacancy they’re going to open, depending on the profit to cost ratio of approaching vacancy versus an unemployment vacancy.

And at that point we’ll have the full blown, you know, picture of what is going on. And we can quantify actually, how strong, you know, the mechanism through poaching vacancies is. And we can make back of the envelope calculations to see by how much can the Fed funds rate go up, without generating unemployment.

Lloyd: That is really fascinating. Whenever you’re done with that part of your research, we’d love to have you back to tell us all about it.

Restrepo-Echavarría: Fantastic. I would love to.

Lloyd: Paulina, thank you so much for joining us today. And if you want to learn more about Paulina’s research or others at the Saint Louis Fed, you can visit stlouisfed.org. A reminder that you can subscribe to timely topics on Apple Podcasts, Spotify, or anywhere you get your podcasts.

I’m Tim Lloyd. And from the Saint Louis Fed, you’ve been listening to Timely Topics.