Related Links

• Multilateral Comovement in a New Keynesian World: A Little Trade Goes a Long Way
• How Does Trade Impact the Way GDP Growth and Inflation Comove Across Countries?

Transcript

Tim Sablik: Hello, I'm Tim Sablik, a senior economics writer at the Richmond Fed. My guest today is Paul Ho, an economist in the Research department of the Richmond Fed. Paul, welcome back to the show.

Paul Ho: Thank, it's great to be here.

Sablik: Policy discussions at the Fed tend to focus most on what's going on in the U.S. economy — GDP growth, inflation, unemployment, or other metrics like that. But Fed officials also pay attention to developments overseas. And over the past three years, we've had no shortage of global shocks and events, starting with the COVID-19 pandemic.

You co-authored a working paper last year, along with fellow Richmond Fed economists Pierre-Daniel Sarte and Felipe Schwartzman, exploring how countries can import and export shocks to inflation and economic growth through the trading of goods and financial instruments. What motivated you to explore this topic and what were you hoping to learn?

Ho: As you said, the events of the past three years have emphasized how much we live in a globally connected world.

You could think of the COVID-19 pandemic as a global shock that affected the whole world. I mean, it directly hit economies everywhere. But we also had reports during the pandemic of events elsewhere in the world affecting the U.S. economy — for example, a factory shutting down in China or Japan driving up the price of cars or other goods in the U.S. That's an indication of the importance of trade.

I think what is perhaps less obvious is how an event like the war in Ukraine might affect the U.S. Ukraine is not a huge country and by no means a major trading partner of the U.S. Yet you saw in multiple FOMC statements, when the Fed announced updates to its monetary policy, some version of the war and related events are creating additional upward pressure on inflation and are weighing on global economic activity.

So, how could something happening in Ukraine have so big an effect on U.S. monetary policy that it warrants a mention in what is a relatively short statement on the economy? With all this as a backdrop, it seemed like a good time to think hard about how countries are connected through trade. We see inflation rising simultaneously in several countries. Is that coming from a shock that hit everyone? Or is that from a shock that hit one country but spilled over to other countries through this trade network?

Economists have thought for a long time about many of these questions about why economic variables co-move across countries. But recent events were really a reminder of how important this multilateral trade network is and how much events in one country can affect the rest of the world.

Sablik: I want to follow up on your point about Ukraine. In your paper, you note that for many large economies, trade accounts for only a small share of purchases, about 20 percent. In the case of the U.S., has this led economists to pay less attention to trade when trying to understand these co-movements in GDP growth and inflation across countries?

Ho: One is tempted to say, well, trade is a relatively small share of GDP, so we aren't losing too much if we ignore it in our models. For some questions, this is very reasonable. As economists, we have to make some hard choices about what ingredients to include or not include in our models.

That said, there is a huge literature thinking about international trade and the sources or implications of trading patterns and other related questions. Where we came in was to figure out how to integrate some of these tools from the international trade literature into quantitative macroeconomic models. This gave us the best of both worlds. We had a tractable way from trade economists to think about the complicated global trade network, but we could also answer questions that macroeconomists care about such as what drives business cycles or what is the effect of monetary policy.

Sablik: Your model features multilateral trade, meaning that countries can trade with multiple partners. How does this affect the way that spillovers work?

Ho: When you have two countries, say the U.S. and China, whatever happens in China spills over to the U.S. and then that's the end of the story. With more countries, you start to realize that rising costs in China don't just affect the U.S. They also affect the EU. When the EU economy is hit, that also spills over to the U.S. And whatever happens in the U.S. then spills over to Japan, and so on and so forth.

In other words, the total effect of the Chinese rise in costs on the U.S. is not just the initial direct effect, but the accumulation of these multiple layers of indirect effects. These can snowball to generate substantial effects. And we've seen this unfold anecdotally over the past three years. There were numerous reports of how shutdowns and other events in foreign countries impacted the U.S. economy, even though trade with these countries may only constitute a relatively small part of the U.S. economy.

Our model tries to resolve this tension and show how you can generate spillovers that are quantitatively important. The key is that you need to account for the accumulation of these direct and indirect effects from the trade network.

Sablik: How did you go about isolating the spillover effects from trade versus the effects from shocks that hit multiple countries at once, like the pandemic?

Ho: In the absence of trade, shocks that hit multiple identical countries at once will just move these different countries in parallel. With empirically observed trade patterns, our theory gives precise predictions about how a shock in one country spills over to other countries, which is going to look different than if we just had global shocks like the pandemic. On top of that, trade changes how these global shocks propagate through the world.

The model takes all of this into account. Then, we bring to the table macroeconomic data like GDP growth and inflation, together with trade patterns and exchange rates, and then use the model to disentangle these country-specific shocks versus global shocks.

The model can also tell us how the correlations would look like if we lived in a world without trade or a world without global shocks. Then, these counterfactuals can give us a sense of how important the spillover effects of trade are.

Sablik: I think we've kept our listeners waiting long enough. What were your key findings in the paper?

Ho: The first question we wanted to ask is, how important is trade? We find that it is indeed quantitatively important on two fronts.

First, trade plays a big role in generating these cross-country correlations. If you lived in a world without global shocks [and] all the co-movement came from trade, you'd still have 90 percent of the cross-country correlations in GDP growth. On the other hand, in a world without trade, the cross-country correlations in GDP growth and inflation would be about a third of what we see in the data.

Next, due to trade, domestic variables can respond substantially to foreign shocks. One example we looked at was the effect of an inflationary shock in the EU on U.S. inflation — think of this as energy prices rising as a result of the Ukraine war. We see that the U.S. inherits 20 percent of the EU response. This is pretty substantial. All these results suggest that trade spillovers are important.

The next question we wanted to ask was, how can trade be this important when it only consists of the small share of GDP? To that end, again, we show that these indirect effects are key. The reason that trade can generate these large correlations and spillovers is that these shocks propagate across the entire trade network. So again, when inflation goes up in Europe, that doesn't just hit U.S. imports from Europe. It also causes inflation to rise in China, which then affects the U.S., which affects Canada, and so on and so forth. It's when these multiple layers of effects add up that we get a large total effect.

Sablik: Economists often invoke the Phillips curve — which is the relationship between inflation and slack in the economy, typically measured by unemployment — as sort of a rough guide for monetary policy. This usually only takes into account domestic measures of inflation and slack. But your research suggests that policymakers may need to think in terms of a global Phillips curve. Can you explain?

Ho: As you said, the Phillips curve describes the relationship between inflation and slack in the economy. When inflation goes down, slack tends to increase. This is important for monetary policy because it says you want to increase interest rates enough to reduce inflation, but not to the point where unemployment goes up too much.

So where does the Phillips curve come from? First, when input costs such as wages go up, firms raise their prices, which increases inflation. But then this is connected to slack because firms are simultaneously responding to domestic demand. So, they cannot arbitrarily raise prices — set too high a price and there won't be enough demand for the firm's output.

This chain of logic breaks down once we have trade. Firm costs no longer depend just on domestic conditions since the inputs come from foreign countries as well. So, if costs in China increase and U.S. manufacturers use Chinese goods as inputs, then it becomes more expensive to manufacture goods in the U.S. and prices in the U.S. will go up. On top of that, firms in the U.S. need to worry both about local and global demand. If Chinese demand for U.S. goods goes down, then U.S. firms are going to need to lower prices or reduce production.

The upshot of all of this is that inflation in the U.S. is not just connected to U.S. real variables, but also to real variables and exchange rates in all other countries, hence a global Phillips curve.

Sablik: Based on your findings, should monetary policymakers be paying more attention to global developments? Or is that something that's already factored into their decision making?

Ho: I think the main message we have for policymakers is that we shouldn't just ignore what happens in certain foreign countries because we think the U.S. doesn't trade much with those countries.

That said, I think external developments are already on the radar, certainly for the Fed. The FOMC statement and meeting minutes are evidence of this.

The big question is, how exactly should we think about this quantitatively? The last thing you want to do is to say trade with this country only constitutes 1 percent of U.S. GDP, so whatever happens there will only change U.S. GDP by some small fraction of 1 percent. How then should we weight different countries? How important are foreign developments compared to domestic events? This is where our paper comes in by giving more structure to think about these problems, particularly as policymakers try to respond in real time.

Sablik: Paul, thanks so much for coming on the show to discuss your research.

Ho: Thanks for having me.

Sablik: Listeners can find a link to the paper we discussed today on the show page. And if you enjoyed this episode, please consider leaving us a rating and review on your favorite podcast app.