Economists widely believe that, absent strategic considerations such as agency problems, financial incentives represent a dominant and effective stimulator of human productive activities. In production settings that are cognitively demanding, however, the effectiveness of financial incentives may be moderated by individual cognitive abilities and motivational characteristics. As a useful metaphor for the moderating channels, Camerer and Hogarth (1999) propose an informal capital-labor-production (KLP) framework, describing how financial incentives may interact in non-trivial ways with intrinsic motivation in stimulating cognitive effort (labor), and how the productivity of cognitive effort may in turn vary across individuals due to their different cognitive abilities (capital). Even if strong financial incentives induce high effort, both financial and cognitive resources may be wasted for individuals whose cognitive constraints inhibit performance improvements. This prediction, if warranted, calls for attention to individual cognitive abilities in designing efficient incentive schemes in firms, experimental settings and elsewhere.
This dissertation examines how financial incentives interact with intrinsic motivation and especially cognitive abilities in determining cognitive performance. In Rydval (2003), I present an initial review of the literature, theoretical issues and outstanding questions pertaining to the KLP framework. I document how the KLP framework has been addressed in economics, psychology and other fields, especially noting the lack of empirical evidence on the interaction between financial incentives and cognitive abilities. Building on the review, Chapter 1 of the dissertation illustrates that general cognitive abilities appear at least as important for performance in a psychometric test as does a sizable variation in piece-rate financial incentives.
Chapter 2 focuses on the interaction between financial incentives and task-specific, as opposed to domain-general, forms of cognitive capital, the role of which is central to the KLP framework of Camerer and Hogarth (1999) and has long been studied in cognitive science and behavioral decision research. Using a task situated in an accounting setting, I show that both financial incentives and task-specific cognitive capital, and especially their interaction, matter for performance. In particular, the effect of task-specific cognitive capital—proxied by accounting knowledge—on performance is stronger under performance-based financial incentives as compared to flat-rate incentives. The interaction effect arises because performance-based financial incentives lead to better performance only for individuals with more task-specific cognitive capital.
Chapters 1 and 2 chronologically precede Chapter 3. They both revisit previously collected experimental datasets that have certain deficiencies. Neither of the datasets offers sufficient information on individual characteristics that would permit accounting for potentially important sample composition differences. The dataset used in Chapter 1 further offers only an endogenous measure of general cognitive abilities with respect to performance in the psychometric test. The results presented in Chapters 1 and 2 should therefore be viewed mainly as illustrations motivating the much more thorough analysis in Chapter 3, which makes the core contribution of the dissertation.
In Chapter 3, I empirically test the key theoretical building block of the KLP framework, namely the causal effect of cognitive capital on performance. Drawing on contemporary cognitive psychology, I measure general cognitive capital by working memory—a robust predictor of general fluid intelligence and performance in tasks requiring controlled information processing. Since pre-existing task-specific cognitive capital (expertise) is vital for performance in many field cognitive tasks but is hard to measure, I intentionally minimize its potential relevance by designing an experiment where working memory arguably becomes the main component of cognitive capital, besides experience acquired endogenously through on-task learning.
Specifically, as a tool for identifying the causal effect of working memory, I design a time-series forecasting task that requires maintaining forecast-relevant information accessible in memory while simultaneously processing it, hence activating precisely the type of cognitive capital that working memory theoretically represents. To identify the causal effect of working memory on forecasting performance, two screens with forecast-relevant information are displayed either concurrently or sequentially. Since the sequential (concurrent) treatment features higher (lower) working memory load, working memory should be a stronger (weaker) determinant of forecasting performance, after controlling for other cognitive and personality (especially motivational) determinants of forecasting performance. This hypothesis is confirmed for individual differences in asymptotic forecasting performance.
I therefore show that the effectiveness of high-powered financial incentives as a stimulator of economic performance can be moderated by individual heterogeneity in cognitive capital in a causal fashion. (Abstract shortened by UMI.)