The US economy surpassed expectations in 2023.

The economy should slow in 2024, and history shows that conditions can deteriorate rapidly. But resilient activity continues to increase the chance of a soft landing, especially if the Fed begins easing soon.

A solid end to the year

US GDP growth was a robust 3.3% annualized in Q4 – the largest upside surprise relative to consensus expectations (2%) since 2019. This partly reflects strong contributions from the volatile inventory and trade components. But, even excluding these, underlying growth was 2.7% annualized.

Consumer spending delivered the bulk of growth, rising 2.8% annualized, powered by faster real income growth and another decline in the savings rate. There are few signs yet of depleted savings, or tight credit conditions, holding back spending. Business fixed investment was up 1.7% annualized, helped by gains in residential, structures, equipment, and intellectual property spending. Non-residential structures spending continues to slow as the boost from IRA and CHIPs legislation fades.

Government consumption and investment delivered another strong quarter, at 3.3% annualized, bringing the gain over 2023 to 4%, the strongest in 20 years. While the fiscal stance was a big tailwind last year, we think it will be a headwind during 2024 as the deficit narrows.

Zooming out, this time last year the Bloomberg consensus for 2023 US GDP growth was 0.5%, with most forecasters, ourselves included, expecting downturns (Chart 1).

Chart 1. The US economy beat expectations in 2023

Source: Haver, abrdn, January 2024.

Few signs of recession on the horizon

Barring enormous revisions, it seems safe to say the US economy is nowhere near a recession. It is officially the National Bureau of Economic Research's (NBER) job to define when a recession starts and ends, with its Business Cycle Dating Committee judging this as a significant decline in activity, spread across the economy, lasting more than a few months. It uses six indicators encompassing household income and spending, employment trends, industrial activity and retail, wholesale, and manufacturing sales to test if these depth, diffusion, and duration benchmarks have been met.

Averaging across all recessions from 1965, these indicators clearly deteriorate at the onset of a downturn, capturing the broad and pronounced downswing in activity taking hold (Chart 2).

Chart 2. The NBER recession metrics are not leading indicators

Source: Haver, abrdn, January 2024.

Interestingly, ahead of a downturn, there is little advance warning in many of these series. Indeed, while growth rates might slow over the 12 months before a recession, the move is not particularly dramatic (Chart 3).

Chart 3. Most NBER indicators slow rather than stall in the lead up to a downturn

Source: Haver, abrdn, January 2024.

The exceptions are industrial production, which is typically a leading cyclical indicator, and manufacturing and trade sales which is likely to capture weak demand in industry, wholesale, and retail. These clearly start to deteriorate ahead of the downturn, with the three-month annualized rate down 0.3% annualized and 0.5% annualized, respectively, on the eve of recession.

Industrial weakness is the main red flag among the NBER indicators

For now, the data remains solid (Chart 4), consistent with the Q4 GDP print and the strong consensus that the US economy is performing well, even if views remain somewhat split over a hard versus soft landing later in 2024.

Chart 4. Most of the NBER indicators look solid

Source: Haver, abrdn, January 2024.

The one exception is the industrial sector, which has been struggling for several quarters and is indeed even weaker than we typically see ahead of a recession. This would typically be a worrying cyclical trend, but the rebalancing of consumer spending away from goods towards services after the pandemic explains much of this sluggishness, and there are few signs of contagion to broader activity.

What should we be looking for?

Some of the more severe recessions in the historical data may not be the best benchmark for a mild business cycle downturn. If the ‘severe’ Volker recessions of 1980–82, the Global Financial Crisis and the pandemic are excluding, the behavior of the NBER indicators looks quite different (Chart 5).

Chart 5. Mild recessions show smaller and less synchronized shifts in activity

Source: Haver, abrdn, January 2024.

Industrial production and real manufacturing and trade sales still provide the earliest indication that trouble is on the way, with these data peaking three to six months before the official start of the recession. By contrast, personal income and spending typically peak around the start of recession, or shortly after, before falling over the subsequent six months. The labor market indicators meanwhile are the slowest to react, stagnating in the early months of a downturn before starting to fall.

The onus is on the data to turn

The strength of US growth through 2023, coming alongside a moderation in inflation, makes a soft landing very possible. This is because progress on lowering inflation allows the Fed to accommodate the easing in financial conditions that has occurred recently, even if they want to push back on the specific start date of rate cuts. The longer the economy remains resilient, the more likely it is that the peak impact from earlier tightening has been and gone.

The economy, however, is not completely out of the woods. Interest rates remain above neutral. Fiscal policy will be a slight headwind to growth. Consumers have run down the bulk of their excess savings. And corporates are having to refinance onto higher interest rates. Moreover, our work on NBER indicators tells us that a downturn can emerge quite quickly and be hard to spot in real time, especially when the downturn is relatively mild.

Taken in the round, our conviction is mainly around the economy slowing during 2024 and the Fed easing policy. Whether or not this constitutes a recession is perhaps 50/50 at this point.