In their 2020 book Radical Uncertainty, John Kay and Mervyn King attack the practice of assigning probabilities to unknowable outcomes. They cite a banker who called the 2008 global financial crisis (GFC) a “25 sigma” (effectively impossible) event, or the military advisers who, after many revisions, assigned a 70% probability of finding Osama bin Laden when raiding his compound in Abbottabad. These events had no precedents, no past data to study, and could not be predicted with anything approaching mathematical certainty. But even if they are not entirely defensible, probabilities help decision-making and preparation for worst-case scenarios.
Economists are often asked to offer a probability of recession in the year ahead. Given the far-reaching consequences of contractions, we do make a good faith effort to respond.
With ample historical data to study, many have attempted to create statistical models of recession risk. Data series abound, from market volatility to industrial production to employment, which may reveal historical correlation with downturns. As an example, yield curve-driven estimates of recession risk are currently low; despite a narrowing of spreads between longer-duration bonds, the full yield curve remains firmly upward-sloping. The limitation of any of these models is that they rely on past relationships, which may not necessarily be repeated in future scenarios.
Estimates are not always well-quantified. Several economic surveys ask their panelists to contribute their estimate of the probability of a pending recession. Respondents provide a number, but they do not need to show their work. Some may have sophisticated models to assign a probability, but others are likely making a simple guess.
In aggregate, surveys of recession risk have started to show a rising trend, likely driven by sentiment and current events. Risks to the outlook are growing. Confronted with headlines about rapid inflation and high oil prices, rising rates, COVID lockdowns and regional wars, it is easy for forecasters to feel bearish and adjust their recession probability upward. These survey-based estimates are rising more quickly than many models are, as most actual data points are do not reflect the uncertainty we feel.
Some of the current discomfort is the unwelcome feeling of slowing down. The U.S. economy grew by 5.7% in 2021, its fastest pace since 1984, a rate more typical of an emerging market. Growth rates were outstanding across most developed markets. As advanced economies return to their more normal trends, it may feel like the onset of a recession. But slowing is not contraction, and slower growth is still growth.
|
Economic indicators are too strong to expect a recession in the U.S. this year. |
