Related Links

• Understanding the K-Shaped Economy
• District Dialogues

Transcript

Tim Sablik: My guest today is Urvi Neelakantan, a senior policy economist at the Richmond Fed. Urvi, welcome back to the show.

Urvi Neelakantan: Tim, it's always great to speak with you about our work.

Sablik: Today, we're going to be discussing the K-shaped economy, which was the topic of a recent District Dialogues event that you participated in here at the Richmond Fed. You presented some research on this topic that you've done with your colleagues here at Richmond, and we'll be delving into some of what you found in that research.

I think it's often helpful to start these conversations with some definitions. Listeners may have heard the term "K-shaped economy" recently. I was checking on Google Trends and it seems that interest in the term increased substantially starting late last year. Even if listeners have heard of it, they may have varying definitions in mind about what this term means. What definition did you and your colleagues use when you started to explore this topic?

Neelakantan: I'm really glad we're kicking this off by defining terms because my sense is the same as yours — that definitions do vary, especially in the public sphere.

In the work that I did with my colleagues Kyle DeMaria, John Jones and Joe Mengedoth, we landed on a pretty strict definition of the K-shaped economy. We say that the economy is K-shaped when economic outcomes for the bottom 20 percent of the household income distribution worsen, while outcomes for the top 20 percent improve. I'll refer to that first group, the bottom 20 percent, as low-income households and the second group, the top 20 percent, as high-income households.

The reason I characterized our definition as strict is because if outcomes for all households improve, but they improve more slowly for low-income households than high-income households, we would not call that a K-shape. It may sound a bit pedantic to say we're not calling it a K because it doesn't really look like one, but I think there's a deeper reason why it matters. One of those reasons is that in a high-income country like the U.S., we probably don't want to see outcomes for low-income households getting worse in absolute terms.

Sablik: Thinking about it — the K-shape — if it's getting worse, they're going down the lower leg of the K versus going up the upper leg. If they're both going up at different speeds, then that would be a strangely shaped K.

Neelakantan: Exactly.

Sablik: Maybe you can talk a little bit more about why might it matter if the economy is K-shaped and why would that be a concern for economic policymakers at the Fed potentially.

Neelakantan: I didn't touch on this during the District Dialogues event, but one set of outcomes that were firmly in my mind were employment and unemployment. In that context, a K-shaped economy would be one where labor market outcomes for higher-income households improve, while labor market outcomes for low-income households get worse.

We've seen this happen. These episodes tend to be brief, but previous research has shown that even these brief K-shapes can have long-run scarring effects. We ourselves have work in progress that suggests that low-income households can get persistently stuck cycling between long-term nonemployment and low-income employment.

Given everything I've just said, you can clearly see how K-shapes might be of concern to policy makers at the Fed, whose mandate includes effectively promoting maximum employment.

Sablik: Based on your definition, K-shape refers to different outcomes for households based on their income groups. Naturally, then, the growth of income for those different groups is going to be key to whether outcomes are K-shaped. How have real incomes grown for these different groups over time?

Neelakantan: Between 1994 and 2024, low-income households saw their average real income grow by 13 percent, middle-income households by roughly 26 percent, and high-income households by about 63 percent. By our strict definition, if we look over this entire three-decade period, real income grew for all three income groups and so we don't describe this entire period as K-shaped.

Having said that, we do see K-shapes in the wake of the recessions that occurred during these three decades. We see that K-shapes are evident in the wake of the 2001 recession, after the 2007-2009 recession. We do not see K-shapes in income after the COVID recession, in part because outcomes for high-income households also briefly got worse. We see the three groups moving in tandem after the COVID recession.

Sablik: So, it wasn't a K because all groups suffered a decline after that period.

Neelakantan: Yes, and then all groups recovered.

Sablik: Do income sources differ between these household groups, and do those differences offer any sort of insights on this K-shaped question?

Neelakantan: Yes, the shares of income that come from different sources do indeed differ by income groups.

One piece of income is net taxes and transfers. The word "net" is important there. If a group pays more in taxes than it receives in transfers from the government, then the net reduces their income. That's what we see for high-income households. For low-income households, the opposite is true. Net taxes and transfers are positive and, in fact, account for more than half of their income.

You asked about whether that offers any insights on the K-shaped question. I don't know that I can answer that directly. But to me, this is suggestive evidence that the tax and transfer system has a role to play in mitigating that lower arm of the K.

Sablik: In thinking about mitigating factors, another thing that you talked about at the District Dialogues event is how might economic mobility complicate the assessment of whether or not the economy is K-shaped or how we interpret a K-shaped economy.

Neelakantan: Economic mobility in the U.S. is low. If you're born in a lower-income household, it's more than likely that you'll remain as part of a low-income household over your life cycle.

That makes the K more concerning. If people moved freely up and down the K, then their income is volatile over their lifetime, but it's also, on average, in the middle. But if you know that once you're born on one arm of the K you're very likely to stay there, then that makes the K-shape more concerning. Even if you're on the K for a little while, that has long-term scarring effects.

Sablik: Another way that people talk about the current economy being K-shaped is in terms of consumption. There have been some arguments since the recovery from the pandemic and the current period that high-income households are consuming at the same rate or even greater rates, while low-income households are pulling back in the face of inflation and rising prices. Did you look at household consumption as well?

Neelakantan: We've talked about income, but income matters for what it can buy you. You're either using it to consume now, or you're saving it to be able to fund consuming something in the future, or you're using it to pay off debt that you took on to consume something in the past. So, we did look at household consumption.

Dispersion of consumption growth over that same 30-year period that we studied was much smaller than dispersion of income growth. Part of it is due to what I just described. Households don't spend exactly what they earn each year. They save and borrow, they smooth their consumption over time. That helps consumption not be quite as volatile as income.

Talking specifically about K-shapes in consumption, there are modest K-shapes in 2002-2004, 2011-2013, 2021-2023. By now you see the pattern — it's after each recession.

The behavior of consumption in 2021-2023 is particularly notable because income, as I mentioned earlier, was not K-shaped over this period but consumption is. And, in contrast to the previous periods, middle-income households' consumption rises rather than falls.

We haven't talked about the middle much yet, but they are also sometimes part of the downward or upward slope of the K. The post-COVID period is one where the consumption of middle-income households actually rises.

Sablik: One thing you mentioned in your District Dialogues presentation is the data that we have on this is lagged and backward-looking. Did you learn anything from the discussions at the District Dialogues event or other sources that can help fill in the data gaps about the economy?

Neelakantan: Let me speak to the data challenge first. The data we used for income is the Current Population Survey. It's a nationally representative survey. It asks people about their income the previous year. So, even though the 2025 survey is out, what we learn are the incomes of people in 2024.

So, what do we do in the meantime? One of the things we do is we look at data sources that come out more quickly. They may be less complete, but these days they tend to be quite representative. The New York Fed is one such data source and they have some posts about the K-shaped economy that we discussed with them.

The other thing that helps us is that there are people from our team who are always doing on-the-ground sensing. They go out into the community. District Dialogues is an example of an event where we bring people from the outside in and they can tell us in real time what they're seeing from talking to consumers and businesses and community members.

Data sources that come out more frequently, combined with our sensing apparatus, can help us fill these gaps. That said, I am not going to venture to tell you what alphabet describes the economy right now.

Sablik: Yeah, that's the big question.

On that note, are you or your colleagues planning to continue or expand your research on this topic?

Neelakantan: John Jones and I have work in progress with two other co-authors, Grey Gordon and Kartik Athreya, looking at binning people into income groups and looking at outcomes over the life cycle for them. Relevant to what we talked about earlier is how much mobility we see across income bins you want to look at.

Sablik: Urvi, thanks for joining me today and sharing with our podcast audience everything that you have learned about the K-shaped economy.