A growing number of studies find a positive link between employee satisfaction and the financial performance of companies. Two examples are Glassdoor’s report Does Company Culture Pay Off? and a recent study by Alex Edmans of the London Business School. But an often-repeated criticism of this research is that it only shows a correlation between happy workers and financial returns, not a causal link.
But a new study published this summer in the Journal of Corporate Finance directly tackles that criticism for the first time. Using a common econometric method known as “instrumental variables,” the authors show a clear causal relationship between employee satisfaction—as measured by Glassdoor company reviews—and subsequent company financial performance.
The new study is “Family Firms, Employee Satisfaction, and Corporate Performance” by four authors from the University of Kansas: Minjie Huang, Pingshu Li, Felix Meschke and James P. Guthrie.
Putting Glassdoor Data to Work
In the study, the authors use a unique data set of more than 100,000 Glassdoor employee reviews between 2008 and 2012. They statistically examine whether company ratings predict two commonly used measures of financial performance: Tobin’s q (which is a measure of a company’s current market value) and return on assets.
The main concern with research linking employee happiness and company performance is what economists call “endogeneity.” Although employee satisfaction might cause better financial performance, the causal link might also run the other direction, with better financial returns leading to higher pay, more lavish perks, and thus higher worker satisfaction. Further complicating things, there may be a third factor that causes both financial performance and employee satisfaction. Economists teach that you can’t untangle this complicated web of causality using ordinary statistical methods.
Untangling Cause and Effect
One common way economists get around this problem is a method known as “instrumental variables” or “two-stage least squares.” It’s a commonly used method in applied economics and allows us to pin down causal links in cases like this where we can’t run carefully controlled, randomized experiments on companies and workers.
The way it works is simple. Suppose we want to estimate the causal impact of employee satisfaction on company financial performance. We tackle the problem in two stages. First we find something that predicts employee satisfaction but is unrelated to financial performance; this is called an “instrument” for worker satisfaction. Using that instrument we generate predicted or “fitted values” for employee satisfaction, giving us clean variation in employee satisfaction that’s not confounded by other factors. In the second stage, we use this “clean” variation in worker satisfaction to estimate the causal link to company financial performance.
In this study, the authors found a clever instrument for employee satisfaction: the fraction of a company’s reviews coming from current versus former employees. Current employees tend to rate employers higher, for the simple reason that most former employees have left their positions due to being dissatisfied—what economists might call “low-quality job matches.” They show the fraction of reviews from current employees predicts company ratings, but is uncorrelated with financial performance, providing an ideal instrument to nail down the causal impact of worker satisfaction on performance.
The Bottom Line
The key findings from the study are in the tables below. In Table 6, the first two columns show the usual correlation between employee satisfaction and the market value of companies—something that’s been studied many times before. Column 3 is what is really original: it shows the causal effect of employee satisfaction on company value based on the “instrumental variables” approach described above.
In Column 3, the authors find a statistically significant causal link between employee satisfaction and the market value of companies. They find that a one-unit change in company rating on Glassdoor’s scale of 1 to 5 has a 0.141 effect on the market value of companies, as measured by Tobins q.
How big is that effect? The average Tobin’s q is 1.79 for the companies they studied. So that’s a 0.141/1.79 = 7.9 percent average jump in company market value from a 1-star increase in employee satisfaction—a huge economic effect.
The authors also looked at the impact of employee satisfaction on a company’s return on assets (ROA). The third column of Table 7 below shows the causal impact. They find a one-unit change in employee rating leads to a statistically significant 0.524 increase in return on investment—further evidence of the positive link between employee satisfaction and a company’s bottom line.
Here’s how the authors summarize their findings: “Consistent with previous studies, our results suggest that corporate culture, as assessed by employees, helps predict subsequent firm performance… our finding provides empirical support for recent theories of the firm that model employees as key corporate assets.”
For more research on the link between employee satisfaction and company stock returns, see our recent study of Glassdoor’s “Best Places to Work” companies.