Why is it so hard to forecast a recession?

By  Manfred Keil and Yao Li | Inland Empire Economic Partnership

In early January, we published an article titled “The recession of 2023: Is it a myth or a reality?” The conclusion was that while there were some indicators that suggested the U.S. economy was in the last 12 months of a recession, these were not sufficiently strong signals to forecast an economic downturn for 2023 with confidence, meaning the probability of it occurring was not zero but relatively low.

We have changed our mind (“twist”) and want to tell you why (“shout”).

We now believe there will be a recession by the end of the year and our forecast is based on the following observed behavior of some leading economic indicators. In doing so, we now join the ranks of others including former Treasury Secretary and Harvard President Larry Summers, who has pushed the story line that there has not been a single episode post World War II in U.S. economic history where the inflation rate was above 4% and the unemployment rate was below 5% without a subsequent recession.

Why is it so difficult to forecast recessions?

Forecasts, in general, are based on past observed behavior of similar situations — just like forecasting a volcano outbreak is based on sensor movements of previous outbreaks. The problem with recessions is that they are fairly rare — there have been 13 downturns for the national economy since 1946. Hence we have to base our forecast, in essence, on a very small sample. Given our new forecast, you would have to argue that “this time is fundamentally different” if you wanted to disagree with our assessment or question our ability to correctly read the output of these sensors..

First, where do we currently stand in terms of output growth? The national economy  experienced two quarters of negative growth at the beginning of 2022. But it is the dating committee of the National Bureau of Economic Research, or NBER, that is the arbiter of recession dates, not the commonly used “two quarters of negative growth.”

And the NBER has not declared a recession, even if they do so with considerable delay as they did in 2008, it is unlikely to declare this episode a recession. This temporary decline in real GDP was followed by two quarters of relatively strong growth of close to 3%, the long-term average growth rate for the U.S. economy, which is also the required growth to keep the unemployment rate from rising.

Note, however, that while much of the fourth-quarter increase in GDP came from consumer expenditures, that within this category, it was almost exclusively the result of consumers spending more money on services, rather than durables or non-durables: people seem to leave home more often to get entertained. But a large chunk of the GDP increase also came from inventory buildup — and businesses will not allow inventory to accumulate further unless they feel consumers are going to buy more (durable) goods. Finally an unusual decline in imports also boosted the numbers, again suggesting U.S. residents spend less. Without these two factors — inventory buildup and import reduction — GDP growth would have been close to only 1%.

What caused us to see the signal turn from green/yellow to red?

First and foremost, it is the continuous and prolonged inversion of the interest rate spread, or, what is called, the inversion of  the yield curve — the difference between interest rates on 10-year government bonds and 3-month treasury bills. When this happened in the past, a recession followed, and there were very few to no false positives.

Next housing starts — that is, new houses being built, not existing houses being sold. The change in housing starts from a year ago is close to a decline of 400,000 units, and the level is below its long-term monthly average of 1.5 million units a month  (annualized).

Another sensor, the average work week in manufacturing, a sector that is particularly sensitive to cyclical fluctuations, continues to decline. The Sahm unemployment index, which is the change of unemployment over the last three months compared to average change from a year ago, is unchanged, but that is often the case before the start of a recession. The only sensor that gives us a signal of no upcoming recession is the change of the consumer confidence index which has improved slightly recently although still is close to a 40-year low.

The Conference Board, which is the umbrella organization of business in the U.S., computes the Index of Leading Economic Indicators, a complicated average of some of the variables we mentioned plus some others which we did not talk about. That index has declined to a sufficiently low level for the Conference Board to bet on an upcoming recession.

Finally, there is a local Inland Empire measure unmentioned by other forecasters. The Inland Empire economy has the attribute of “First In, Last Out” when it comes to recessions. The analogy is of a lake that freezes from the periphery (Inland Empire) while the middle of the lake (Greater Los Angeles area) still remains ice free for a while. Estimates of the unemployment rate for the Inland Empire are only provided non-seasonally adjusted by the California Employment Development Department. Even these numbers show a slight increase from 3.4% to 3.6% compared to May 2022. However, using standard statistical techniques to remove seasonal fluctuations from the raw numbers show a recent jump in the Inland Empire unemployment rate to 4.9%, which is a flashing warning light.

Bottom line, we need to shout out that we are now forecasting a recession for the national economy. Time to fasten the seat belt, there is a traffic jam ahead.

We did not talk about the severity of the recession, which is the topic for another article. However, we don’t expect this to be a very bumpy ride — instead we expect a relatively mild recession.

Manfred Keil is chief economist, Inland Empire Economic Partnership, associate director, Lowe Institute of Political  Economy, Robert Day School of Economics and Finance, Claremont McKenna College.

Yao Li is a Ph.D. student, Kellogg School of Management, Northwestern University.

The Inland Empire Economic Partnership’s mission is to help create a regional voice for business and quality of life in Riverside and San Bernardino counties. Its membership includes organizations in the private and public sector.

Older Post Newer Post