“People and Process, Suits and Innovators: the Role of Individuals in Firm Performance”, 2012-09-01 (; similar):
Performance differences between firms are generally attributed to organizational factors rather than to differences among the individuals who make up firms. As a result, little is known about the part that individual firm members play in explaining the variance in performance among firms.
This paper employs a multiple membership cross-classified multilevel model to test the degree to which organizational or individual factors explain firm performance. The analysis also examines whether individual differences among middle managers or innovators best explain firm performance variation.
The results indicate that variation among individuals matter far more in organizational performance than is generally assumed. Further, variation among middle managers has a particularly large impact on firm performance, much larger than that of those individuals who are assigned innovative roles.
…The full MobyGames dataset on the PC games industry covers 25 years 1981–25200618ya and contains 5,794 games with full credits and normalized titles. As will be discussed, the data are further matched with revenue information. Since performance data was limited to commercial games sold 1994–12200618ya, this culled the sample somewhat: 1,970 credited games had revenue information. These games involved a substantial number of individuals in the development process. Core team sizes ranged 1–395, with a mean of 52 people in the core team for games that have both credits and performance information.
In order to differentiate between firm and individual effects, the analysis includes designers and producers who worked on more than one game, and who worked with other combinations of designers and producers rather than repeatedly being part of the same team at the same company. Dropping games with individuals that did not meet those criteria resulted in a final sample of 854 games using revenue information, accounting for just over $5.5$42012 billion of revenue. This ultimately allowed me to incorporate 537 individual producers, 739 individual designers, and 395 companies in the revenue model.
…The analysis shows that behind the veil of the firm, variation in individual managers and innovators has both a large and a statistically-significant effect on the success of individual projects. The impact of producers—the mid-level project managers—is especially high. Individual producers account for 22.3% of the variation in revenue, after accounting for game-level predictors. Individual designers, perhaps surprisingly, had only a marginally statistically-significant impact on revenue, explaining 7.4% of variation. 4 In total, the individuals in just these 2 roles accounted for 29.7% of the variation for the products for which they were responsible. Additionally, the individuals with the managerial role of producer explained more of the variation in performance than the individuals who filled the innovative role of designer.
Firms are also statistically-significant, though they explain slightly less variation, 21.3%, than do individual producers. Additionally, the variation explained at the firm level likely overstates the importance of organizational-level processes relative to individuals because they likely incorporate some of the impact of people whose names and job descriptions do not appear in the credits, such as marketers and company leaders, in addition to other factors that may have been left out of the controls. While some variations in revenue are, of course, attributable to firm-level effects directly, the variations in the performance of individuals for these 2 roles alone is at least as important.