In section 2.2.1 of the target article, Boyer claims that norms and laws cannot adequately account for ownership intuitions. We agree, but think it would be unwise to dismiss norm-based accounts entirely. In many cases, intuitions will conflict with norms and laws, but in others, intuitions may hew to norms and laws almost perfectly, leaving little room for P(Agent, thing) and Brep[Min(A)] representations to perform additional causal work. We think laws that give governments taxation powers provide an interesting case study.
We recently explored this case study in an experiment with 200 Prolific workers (M age = 36.8, 48% female). In the first phase, participants learned that United States homeowners are subjected to a federal capital gains tax when they sell their homes. Next, participants read a vignette about a couple who owned a house and were planning to sell it. The neighborhood where the house was located had greatly improved during their ownership tenure; consequently, the couple was set to make $100,000 more in profit than they would have otherwise.
To manipulate one of the input representations that P(Agent, thing) inherits – “changes in thing” – we randomly manipulated participants' beliefs about the cause for the neighborhood improvement: Some subjects learned that the owners invested their personal time and resources into improving the neighborhood (Owners), others learned that wealthy neighbors had invested time and resources (Neighbors), and still others learned that the federal government had invested time and resources (Government). Subjects in a control condition received no such information (Control). Clearly, the homeowners were responsible for the additional profit (i.e., they improved the neighborhood) only in the “Owners” condition. We then asked participants to provide their opinions about: (1) The degree to which the couple was “responsible” for the additional profit, (2) the degree to which the couple was “entitled” to the additional profit, (3) the percentage of additional profit that “rightfully belongs” to the owners, and (4) the percentage of additional profit the government should receive in taxes. As a reference point, we informed participants that Americans generally pay 25–30% of their earned income in taxes.
Our first measure, perceived responsibility of the couple, confirmed that participants believed that the input of the homeowners' labor increased their “responsibility” for the additional profit. Participants who were told the owners had invested time and effort in the neighborhood reported significantly higher levels of owner responsibility for the profit than did subjects in any other condition (all pairwise comparisons with Owners condition p < 0.001; effect size for the difference between the Owners condition and all other conditions combined was Cohen's d = 0.99). No other pairwise comparison (e.g., Neighbors vs. Government conditions) was statistically significant. This result confirms that people attributed the increased profits to the owners when it was the owners who had in fact improved the neighborhood.
Despite the large effect on perceived responsibility, however, none of the four groups differed on the three other variables. Specifically, our manipulation of responsibility had no effect on whether participants believed that windfall “rightfully belong[ed]” to the owners (p = 0.49), the percentage of the windfall the owners ought to keep (p = 0.42), or the degree to which the owners were “entitled” to that windfall (p = 0.58). Given the important role Boyer gives to “changes in thing” as an input to P() representations, this set of findings is surprising.
Two explanations seem promising. Our first clues are the mean scores reported on the two measures that used percentage scales, “percent owners should keep after tax” (M = 78.7%) and “percent rightfully belongs to owner” (M = 73.2%). Given that we provided “25–30%” as a reference point for what the typical American pays in tax, these values are surprisingly close to what one would expect if subjects merely subtracted the typical rate of taxation (25–30%) from the total profit (100%). That is, our participants' estimates of what ought to occur for distributing this windfall were nearly identical to what tax law specifies (and enforces), and seemed unaffected by which agent's “changes in thing” generated the windfall: Perhaps subjects simply acknowledged the fact that the government generally takes 25–30% of earned income, combined this fact with a pre-existing belief that “other people must follow the law,” and then concluded that the homeowners did not rightfully own, and thus were not entitled, to all of the profit, even though they were causally responsible for all of it.
Here is an alternative interpretation that Boyer might prefer. Perhaps people construe governments as agents – a not unreasonable surmise since people conceptualize governments as entities that possess psychological attributes such as the capacity to know, predict, forbid, allow, encourage, and negotiate – which in other work Boyer (Reference Boyer1996) has argued are the types of psychological properties that most fundamentally lead to perceptions of agency. If so, our participants might have construed our questions about ownership of the homeowners' profits as questions about contested ownership between two agents, and not just as questions about how we infer norms and determine how we should respond to them. Drawing on Boyer's section 8.3, if our subjects perceived the government not only as an agent, but also as an agent with whom they themselves were in a cooperative relationship – that is, they possessed high values of Brep[Min(A)] for the government – then they might have interpreted the government's ability to appropriate a share of the profits as evidence of the government's legitimate ownership.
We regret that we didn't collect data to adjudicate between an explanation based on a psychology of norm adherence and an explanation based on Boyer's account . We might have, for example, manipulated whether subjects viewed the government as a friend or foe. Although this seems to us like a useful direction for future work, for now we will be content with reading Boyer's reactions to the ideas we have raised here.
You have
Access
In section 2.2.1 of the target article, Boyer claims that norms and laws cannot adequately account for ownership intuitions. We agree, but think it would be unwise to dismiss norm-based accounts entirely. In many cases, intuitions will conflict with norms and laws, but in others, intuitions may hew to norms and laws almost perfectly, leaving little room for P(Agent, thing) and Brep[Min(A)] representations to perform additional causal work. We think laws that give governments taxation powers provide an interesting case study.
We recently explored this case study in an experiment with 200 Prolific workers (M age = 36.8, 48% female). In the first phase, participants learned that United States homeowners are subjected to a federal capital gains tax when they sell their homes. Next, participants read a vignette about a couple who owned a house and were planning to sell it. The neighborhood where the house was located had greatly improved during their ownership tenure; consequently, the couple was set to make $100,000 more in profit than they would have otherwise.
To manipulate one of the input representations that P(Agent, thing) inherits – “changes in thing” – we randomly manipulated participants' beliefs about the cause for the neighborhood improvement: Some subjects learned that the owners invested their personal time and resources into improving the neighborhood (Owners), others learned that wealthy neighbors had invested time and resources (Neighbors), and still others learned that the federal government had invested time and resources (Government). Subjects in a control condition received no such information (Control). Clearly, the homeowners were responsible for the additional profit (i.e., they improved the neighborhood) only in the “Owners” condition. We then asked participants to provide their opinions about: (1) The degree to which the couple was “responsible” for the additional profit, (2) the degree to which the couple was “entitled” to the additional profit, (3) the percentage of additional profit that “rightfully belongs” to the owners, and (4) the percentage of additional profit the government should receive in taxes. As a reference point, we informed participants that Americans generally pay 25–30% of their earned income in taxes.
Our first measure, perceived responsibility of the couple, confirmed that participants believed that the input of the homeowners' labor increased their “responsibility” for the additional profit. Participants who were told the owners had invested time and effort in the neighborhood reported significantly higher levels of owner responsibility for the profit than did subjects in any other condition (all pairwise comparisons with Owners condition p < 0.001; effect size for the difference between the Owners condition and all other conditions combined was Cohen's d = 0.99). No other pairwise comparison (e.g., Neighbors vs. Government conditions) was statistically significant. This result confirms that people attributed the increased profits to the owners when it was the owners who had in fact improved the neighborhood.
Despite the large effect on perceived responsibility, however, none of the four groups differed on the three other variables. Specifically, our manipulation of responsibility had no effect on whether participants believed that windfall “rightfully belong[ed]” to the owners (p = 0.49), the percentage of the windfall the owners ought to keep (p = 0.42), or the degree to which the owners were “entitled” to that windfall (p = 0.58). Given the important role Boyer gives to “changes in thing” as an input to P() representations, this set of findings is surprising.
Two explanations seem promising. Our first clues are the mean scores reported on the two measures that used percentage scales, “percent owners should keep after tax” (M = 78.7%) and “percent rightfully belongs to owner” (M = 73.2%). Given that we provided “25–30%” as a reference point for what the typical American pays in tax, these values are surprisingly close to what one would expect if subjects merely subtracted the typical rate of taxation (25–30%) from the total profit (100%). That is, our participants' estimates of what ought to occur for distributing this windfall were nearly identical to what tax law specifies (and enforces), and seemed unaffected by which agent's “changes in thing” generated the windfall: Perhaps subjects simply acknowledged the fact that the government generally takes 25–30% of earned income, combined this fact with a pre-existing belief that “other people must follow the law,” and then concluded that the homeowners did not rightfully own, and thus were not entitled, to all of the profit, even though they were causally responsible for all of it.
Here is an alternative interpretation that Boyer might prefer. Perhaps people construe governments as agents – a not unreasonable surmise since people conceptualize governments as entities that possess psychological attributes such as the capacity to know, predict, forbid, allow, encourage, and negotiate – which in other work Boyer (Reference Boyer1996) has argued are the types of psychological properties that most fundamentally lead to perceptions of agency. If so, our participants might have construed our questions about ownership of the homeowners' profits as questions about contested ownership between two agents, and not just as questions about how we infer norms and determine how we should respond to them. Drawing on Boyer's section 8.3, if our subjects perceived the government not only as an agent, but also as an agent with whom they themselves were in a cooperative relationship – that is, they possessed high values of Brep[Min(A)] for the government – then they might have interpreted the government's ability to appropriate a share of the profits as evidence of the government's legitimate ownership.
We regret that we didn't collect data to adjudicate between an explanation based on a psychology of norm adherence and an explanation based on Boyer's account . We might have, for example, manipulated whether subjects viewed the government as a friend or foe. Although this seems to us like a useful direction for future work, for now we will be content with reading Boyer's reactions to the ideas we have raised here.
Financial support
This research received no specific grant from any funding agency, commercial, or not-for-profit sectors.
Competing interest
None.