More driven? Experimental evidence on differences in cognitive effort by social origin
While the association between socioeconomic status (SES) and test scores is well-established, little is known about social inequalities in effort, which are likely a key explanatory variable in SES performance gaps.
More recent theoretical approaches tend to suppose that higher SES kids will exert more effort. Kohn’s seminal theory of parental values, similarly to Lareau’s more recent account, posits that high-SES parents instil values such as self-determination and autonomy in their offspring. However, to this day, working class occupations tend to involve routine, often manually-intensive, labour. Hence, low-SES children may be socialized into a disposition to work hard, especially when given external incentives such as money or other rewards. This disposition may be masked by the fact that high-SES children are socialized to find academic work more inherently rewarding and less effortful, making analyses based on academic, or academic-style, tests inappropriate. Furthermore, incentive conditions are likely to be a crucial moderating variable, as intrinsic versus extrinsic motivation likely varies by socioeconomic status.
While survey research draws on self-reports and generally does not incorporate incentive conditions, laboratory studies - which may vary incentives - tend to rely on small convenience samples. We combine the strengths of both approaches with large-scale laboratory data on effort collected on over 1,300 fifth-grade students in Madrid and Berlin. We examine multiple measures of real effort under varying incentive conditions to address the following research questions:
1. To what extent do children’s effort levels differ by parental socioeconomic background?
2. Can existing social disparities in effort be explained by differences in susceptibility to incentives?
Cognitive effort was measured in three computerized real-effort tasks, which are designed to demand minimal cognitive ability but nevertheless require sustained cognitive effort. Each task focused on a different cognitive function: the Slider Task on information processing and goal maintenance; the Simon Task on attentional control and inhibition; and the AX Continuous Performance Task on switching between proactive and reactive control.
Pupils performed the tasks under three different conditions. In the “intrinsic” condition, no material incentives were provided. In the “extrinsic” condition, children could win points which were convertible into toys. And in the “tournament” condition, the top three performers were publicly recognized (while material incentives remained in place). The intrinsic condition measured intrinsic motivation; the extrinsic condition measured responsiveness to material incentives; and the tournament condition measured responsiveness to status incentives. To increase external validity through simulating the opportunity costs which characterize effort choices in the real world, pupils chose between performing the task and playing a game in each of the 10 rounds.
Overall, we find a positive association between parental SES and children’s exerted effort on the tasks (see Figure 1 attached). However, the effect seems to vary across incentive conditions. The gap is largest in the intrinsic condition, and shrinks when effort is extrinsically motivated. Hence, there seem to be important variations in how low- and high-SES individuals respond to incentives. The results are summarized in Figure 1 (attached), for different incentive conditions and socioeconomic background indicators.
More recent theoretical approaches tend to suppose that higher SES kids will exert more effort. Kohn’s seminal theory of parental values, similarly to Lareau’s more recent account, posits that high-SES parents instil values such as self-determination and autonomy in their offspring. However, to this day, working class occupations tend to involve routine, often manually-intensive, labour. Hence, low-SES children may be socialized into a disposition to work hard, especially when given external incentives such as money or other rewards. This disposition may be masked by the fact that high-SES children are socialized to find academic work more inherently rewarding and less effortful, making analyses based on academic, or academic-style, tests inappropriate. Furthermore, incentive conditions are likely to be a crucial moderating variable, as intrinsic versus extrinsic motivation likely varies by socioeconomic status.
While survey research draws on self-reports and generally does not incorporate incentive conditions, laboratory studies - which may vary incentives - tend to rely on small convenience samples. We combine the strengths of both approaches with large-scale laboratory data on effort collected on over 1,300 fifth-grade students in Madrid and Berlin. We examine multiple measures of real effort under varying incentive conditions to address the following research questions:
1. To what extent do children’s effort levels differ by parental socioeconomic background?
2. Can existing social disparities in effort be explained by differences in susceptibility to incentives?
Cognitive effort was measured in three computerized real-effort tasks, which are designed to demand minimal cognitive ability but nevertheless require sustained cognitive effort. Each task focused on a different cognitive function: the Slider Task on information processing and goal maintenance; the Simon Task on attentional control and inhibition; and the AX Continuous Performance Task on switching between proactive and reactive control.
Pupils performed the tasks under three different conditions. In the “intrinsic” condition, no material incentives were provided. In the “extrinsic” condition, children could win points which were convertible into toys. And in the “tournament” condition, the top three performers were publicly recognized (while material incentives remained in place). The intrinsic condition measured intrinsic motivation; the extrinsic condition measured responsiveness to material incentives; and the tournament condition measured responsiveness to status incentives. To increase external validity through simulating the opportunity costs which characterize effort choices in the real world, pupils chose between performing the task and playing a game in each of the 10 rounds.
Overall, we find a positive association between parental SES and children’s exerted effort on the tasks (see Figure 1 attached). However, the effect seems to vary across incentive conditions. The gap is largest in the intrinsic condition, and shrinks when effort is extrinsically motivated. Hence, there seem to be important variations in how low- and high-SES individuals respond to incentives. The results are summarized in Figure 1 (attached), for different incentive conditions and socioeconomic background indicators.