In a Nutshell: In macroeconomics, the Phillips Curve claims an inverse relationship between unemployment and inflation: When inflation increases, unemployment decreases and vice versa. But empirical data doesn’t bear it out unless you examine it granularly — as did a recent team that included Allan Timmermann of UC San Diego’s Rady School of Management. In a May 2023 working paper, Timmermann and colleagues found that in a hot labor market, such as the one that exists in the US, the Phillips Curve is very steep, while in cooler markets, the curve flattens. This instability poses a challenge to the Federal Reserve’s strategy of raising interest rates to lower inflation to the Fed’s traditional 2% target while producing a recession-free “soft landing.” Exploiting a trade-off that isn’t even stable and whose shape changes over time is a tough job so Timmermann suggests that a sustainable reduction in inflation to the Fed’s 2% target without triggering a recession will prove challenging.
The Phillips Curve is a macroeconomic theory that stipulates a tradeoff between the inflation rate and how close the economy is to operating at capacity.
The idea is that there’s a natural unemployment rate, and as long as unemployment holds at that level, inflation will neither accelerate nor decelerate. The central bankers at the Federal Reserve hope the resulting equilibrium will achieve an inflation rate not far from the Fed’s preferred target of 2%.
On the other hand, if the labor market overheats so that unemployment decreases, the Phillips Curve expects inflation to shoot up due to a wage-price spiral. That’s when rising wages increase general business costs as the demand for more goods and services increases, and these costs are passed on to the consumer as higher prices.
The theory calls on the Fed to raise interest rates to reduce labor demand and set things working in reverse. But a recession might result if the Fed overshoots.
It makes intuitive sense, but the broadest look at the available data doesn’t bear it out, to the point that many macroeconomists (and politicians) feel comfortable calling on the Fed to be more aggressive in fighting inflation without fearing a downturn.
But now a team, including Allan Timmermann of the University of California San Diego’s Rady School of Management, has taken a creative new look at Phillips Curve data using a granular sector-by-sector approach and found the model highly valid in the post-COVID US economy.
Timmermann and his colleagues, Jonathan H. Wright of the Johns Hopkins University Department of Economics and Simon Smith of the Board of Governors of the Federal Reserve System, suggest the Fed needs to respect the implications of the Phillips Curve and tread lightly in fighting inflation.
“When unemployment is very low, but the economy is running really hot, we see a steep tradeoff between unemployment and inflation,” Timmermann said. “This is in a sense good news as it suggests that even small increases in unemployment from the current low rate can help reduce inflation quite significantly.
“The greater problem is that, as unemployment rises further and the economy cools down, the Phillips curve flattens, and it becomes increasingly difficult to further reduce inflation. The instability of the Phillips curve trade-off also raises the chance of a policy error triggering a recession.”
Finding the Phillips Curve in a Hot Labor Market
Timmermann earned his PhD at Cambridge in 1992 and is the Dr. Harry M. Markowitz Endowed Chair in Finance and Investing and a Distinguished Professor of Finance at UC San Diego’s Rady School.
His research centers on what might broadly be called financial forecasting models — how to use various datasets to forecast everything from asset and market performance to macroeconomic variables. Timmermann’s research also assesses the extent to which economic events are predictable and determines which sources of information to rely on — or perhaps not to rely on — in a given scenario.
“Lately, I’ve been trying to answer the question of who’s more accurate, economic forecasters or weather forecasters,” Timmermann said. “There isn’t a definitive answer because it very much depends.”
That level of skepticism of economics undoubtedly helped Timmermann and his colleagues as they delved into the legendary Phillips Curve. It turns out the real economy is considerably more nuanced than the intuitive one. Perhaps that was an easy assumption as economists observed inflation’s quick rise as labor shortages (and wage-rate increases) materialized in COVID’s wake.
Timmermann and his colleagues observed much of the post-COVID effect emanating from the energy sector as transportation and food costs and wages continued to increase at 4.6% year on year. It added up to a definite relationship between low unemployment and high inflation.
The working paper uses panel regression analysis to comprehensively analyze historical data covering US cities, states, and industries, and a large set of European countries. This statistical method allowed the researchers to explore pooled cross-sections of data at the product category level, across political jurisdictions, and incorporating wages and inflation.
“When you build up these large panels of data in that way, we find ample evidence of these unemployment and inflation tradeoffs,” Timmermann said. “It’s just that they have changed over time.”
Theory Versus Practice in Macroeconomics
Skepticism regarding the Phillips Curve encourages proponents of an aggressive Fed anti-inflation approach to call for continued interest rate increases without much concern for the so-called inflationary “soft landing” where labor demand and ballooning wages gradually recede and the system returns naturally to normalcy.
The work of Timmermann and his colleagues stands not just as a caution to the Fed regarding the limits of its ability to achieve a soft landing but also as a primer on the relationship between rationalistic macroeconomic theory and the real world.
Timmermann’s 2016 book, “Economic Forecasting,” co-authored with Graham Elliott and published by Princeton University Press, assesses the validity of various statistical approaches as they apply to macroeconomic forecasting problems.
In the case we’re discussing here, panel regression analysis revealed a relatively flat Phillips Curve before the first oil price shocks in 1972-73, implying little relationship between unemployment and inflation. That was followed by a gradual steepening until China’s 2001 entry in the World Trade Organization, which signaled a formalization of globalization in the world economy. The curve flattened again.
“That’s because Western economies could suddenly grow faster without generating inflation,” Timmermann said. “They did that for years because they could just import goods from China without elevating the pressures on the labor markets that otherwise would produce those things in places such as Germany and the US.”
COVID caused the curve to steepen again. As the US recovers in a tight labor market with relatively low unemployment and strong wage growth, businesses in tight markets with tight supplies find they have more pricing power. In other words, they can raise prices on products and services without impacting demand.
That’s where we’re at today. Timmermann and his colleagues calculate that they can explain half of the current rise in inflation due to this joint effect of an increase in the natural rate of unemployment and the steeper Phillips Curve.
“It relies on this estimate of the natural unemployment rate, this labor market capacity of the economy, also having increased,” he said.
The Intricacies of the Soft Landing
In the soft landing scenario, inflation gradually slows down to the Fed’s 2% level within the next few quarters without unemployment having to go up significantly. While the Fed’s pain threshold for unemployment is yet to be seen, Timmermann thinks a rate much above 5-6% will cause some hesitation among Fed officials for further interest rate rises.
“Is that achievable? The signs up to this point have been fairly promising in the sense that unemployment has not shot up by that much; we’re still under 4%,” he said.
But Timmermann said that historically, unemployment can, and has, quickly started upward. The critical issue is whether persistent wage increases — an overarching symptom of the hot economy — are consistent with 2% inflation. Wage pressures may gradually decrease if inflation stays below 3-4%.
Meanwhile, job creation remains highly robust, suggesting much recovery energy remains in the economy. The question is whether the Fed will be happy with 3% inflation in this overheated economy.
“The problem is that if you have an average rate of 3%, inflation fluctuates over time,” Timmermann said. “That means that in some years it’s 1% and in others 4-5%, and I don’t know whether that’s something the Fed would necessarily be happy with.”
So far, the Fed has declared its intent to pursue interest rate pressure to move the inflation rate down to where it prefers. Given the conclusions in the working paper, Timmermann said he hopes caution is the watchword.
With mortgage rates so high that many are locked into current housing and disincentivized to buy new homes, and venture capital pricing itself out of the market, there’s something to be said for interest rate moderation.
“The relationship is volatile, and it’s not necessarily a trade-off you want to work to exploit,” Timmermann said.