As the saying goes, “You only get one chance to make a first impression.” The importance of making a good first impression with your employer is well established in conventional wisdom. But what about second and third impressions? In our recent paper, “Do first impressions last? The impact of initial assessments and subsequent performance on promotion decisions,” available here, we examine how readily employers revise their initial opinion about a worker’s quality and potential after observing actual on-the-job performance. We find that managers continue to make promotion decisions in part based on assessments that were made as many as six years before, and long after the assessments are no longer relevant for predicting workers’ future performance.

How quickly employers learn about employee quality is an important issue, and not just for newly hired employees eager to make a good first impression. The speed and accuracy with which firms update beliefs about employee ability help determine how well human capital resources are allocated, which is critical for firm success. Also, how readily employers revise their initial beliefs about worker ability after observing performance is important for providing employees’ motivation to work hard. If employers quickly incorporate observed worker performance into perceptions of worker ability, which in turn affect job assignment and salary decisions, the payoffs for hard work are relatively high. However, if employers are slow to update their beliefs and continue to hold on to initial impressions about workers, incentives to work hard or gain skills are reduced.

To this point, though, evidence on this topic has been limited because researchers generally do not observe firms’ initial beliefs about employees. To overcome this challenge, we take advantage of a unique institutional feature of Minor League Baseball (MiLB). In this industry, workers usually enter the labor market via a formal draft system. In the draft, each team subjectively ranks prospective players according to expected ability, and teams have successive opportunities to select their highest-ranked player from the available remaining labor pool. Thus, each player’s draft position provides a measure of the initial assessment (i.e., the first impression) of ability by his employer, and we are able to compare the effect of this initial assessment to the effect of on-the-job performance for subsequent promotion decisions as players move through the job-level hierarchy within MiLB.

While many features of the workers and job in our setting are perhaps unique to professional sports, we emphasize that the purpose of our study is to examine inputs to promotion decisions, not the performance or behavior of the athletes themselves. We suggest that the process of forming initial beliefs about worker quality, changing beliefs over time based on observed performance, and promoting employees into a job hierarchy is sufficiently general to apply to a wide range of industries and organizations.

Employers form initial expectations about workers’ ability based on whatever limited information is available when workers enter the labor market. This may include where the employee went to school, how they did on a standardized test, or how well they performed in an interview. Over time, firms can update beliefs about employees’ ability by observing on-the-job performance. Prior research shows that how quickly a worker is promoted initially is positively related to subsequent promotion rates. This result is usually interpreted as evidence for firms quickly learning about worker ability and establishing “fast tracks” for better workers to be promoted earlier. An alternative possibility, however, is that managers are susceptible to confirmation bias, and workers are placed on “fast tracks” before they start working and are promoted more quickly due to biased performance evaluations and not due to superior performance. In our paper, we examine whether, consistent with economic theory, the importance for promotion decisions of initial assessments and on-the-job performance reflects information about future ability or whether, consistent with confirmation bias, managers put too much weight on initial assessments.

We analyze promotion and performance data for over 10,000 unique players in 30 different MLB organizations over the period 1987-2014. There is strong evidence that employers’ initial assessments about employee ability predict promotion decisions, even after several years on the job. These differences in promotion rates for players drafted earlier vs. later lead to dramatic differences in career attainment: Employees initially assessed to be in the top quartile of their cohort are nearly six times more likely to reach the highest job level (Major League Baseball) compared with players assessed to be in the bottom quartile. Of course, the fact that promotion rates are related to draft position (even several years later) could just reflect better players being drafted earlier, and these better players are promoted based on superior performance. Based on the data we examined, however, this doesn’t appear to be the case.

Using regression analysis, we find that, after controlling for observed on-the-job performance, draft position is not predictive of future performance. That is, once you take into account how a player has actually performed on the field, draft position is no longer a useful predictor of how he’ll perform in the future. (We do find, though, that draft position does strongly predict how a player will do in their first year, so teams are generally capable of identifying talent in the draft.) However, even in the face of readily available objective performance data (which strongly predict future performance), teams still appear to attach importance to draft position, even though it is no longer useful for predicting future performance (and therefore economic theory suggests it should not be used in promotion decisions).

We also examine how the relative importance of initial assessments and on-the-job performance changes over a player’s minor-league career. While current performance predicts future performance for every year of player experience, draft position does not. For promotion decisions, teams put less weight on draft position compared with actual performance as players gain experience, suggesting that managers recognize that draft position is less useful as time goes on. However, draft position still receives significant weight in promotion decisions in every experience partition, even up to six years after the draft assessment was made. We see a similar pattern when we estimate our promotion and performance models within each job level (i.e., Rookie ball, A, AA, AAA). So, even though teams can easily observe on-the-job performance, and even though players’ ability likely changes substantially over time (players can improve skills, or may get hurt and regress), teams still base promotion decisions in part on their first assessments, which were made prior to when the player even started working in professional baseball, and which are no longer informative about the player’s quality. Overall, our results are consistent with managers exhibiting confirmation bias in promotion decisions by over-relying on their initial assessments of ability compared with observed performance.

Our research highlights the difficulty of properly evaluating employees and promoting the right workers to positions where they can best help the organization succeed. Even in job environments (like the professional baseball setting we study) where objective performance measures are readily available and extensive attention and resources are devoted to performance evaluation and job assignment decisions, managers may continue to rely on outdated first impressions, which can lead to the wrong people getting promoted and in turn, result in reduced employee motivation. Managers should be aware of the potential for biases to impede employee evaluation. We encourage firms to carefully scrutinize their performance evaluation processes and take steps to mitigate bias.

This post comes to us from Professor Dirk Black at Dartmouth College’s Tuck School of Business and Professor Marshall Vance at the University of Southern California’s Marshall School of Business, Leventhal School of Accounting. It is based on their paper, “Do First Impressions Last? The Impact of Initial Assessments and Subsequent Performance on Promotion Decisions,” available here.