Investing.com — Hiring among tech companies has slowed down 5pp/per year since 2022, compared to the historical trend, with about half of this slowdown explained by tech firms overhiring in the 2020-2022 period, Goldman Sachs said in a note.

According to the firm, three overarching factors have been widely blamed as contributing to the hiring headwinds: “hawkish Fed pivot that slowed growth and raised rates, AI efficiency gains, and a correction for pandemic-era overhiring.”

However, the role of each factor is complicated to tease apart.

The firm finds little evidence to support the view that higher interest rates are driving the hiring slowdown. In fact, firm analysts note, “Hiring trends are virtually identical across tech companies that were more and less exposed to higher rates.”

There has been some AI-driven hiring slowdown, but Goldman Sachs believes it has had a small impact. Differences in occupational AI exposure can only account for about half a percentage point of the slowdown in annual tech employment growth since 2022, the firm says.

The firm also notes that AI-layoff announcements appear credible given that the companies that announced AI-related layoffs actually reduced headcounts in AI-related occupations more than companies citing other reasons.

Goldman Sachs finds strong evidence that companies that overhired between 2020-2022 are the ones that are underperforming on hiring. The analysts observed up to 2pp of the slowdown in annual tech employment growth since 2022 is readily explained by headcount normalisation.

While the firm’s analysis does not identify a specific shock that drove the overall slowdown in tech hiring, it does suggest “aggregate factors and a broad sectoral slowdown in hiring explain a large share of the labor market underperformance in recent years.”

To test if high interest rates drove the hiring slowdown, the firm split its universe of public U.S. tech companies based on changes in their interest coverage ratio over the hiking cycle. The analysts find essentially no difference in headcount evolution between the top and bottom ICR change terciles.

Companies in the highest and lowest ICR terciles are hiring no differently from other companies in the sector as a whole.

As the companies that overhired are now hiring less and the companies that claimed to be cutting headcounts for AI-related occupations are the same ones underperforming on hiring, AI-washing is not a widely applicable explanation.

Goldman Sachs’ research finds that AI and post-pandemic headcount normalisation are both responsible for the hiring slowdown, but headcount normalisation is three to four times more important as an explanatory factor.