The Regulatory Policy Institute Research Group

Immigration and the Ultimatum Game

Alongside the Prisoners Dilemma, study of the Ultimatum Game (UG) and its variants is a rich source of experimental observations on human attitudes and conduct which are, or at least should be, of some relevance in the much wider context of public policy development.

The structure of the UG is as follows. There are two parties, anonymous to each other. One, the proposer (P), is given stewardship of a pot of money (a “pie”) and asked to propose/offer a division of that pie to the other, the responder (R). If R accepts, the pie is divided as proposed by P. If R rejects the proposal/offer, neither of the parties gets any of the pie.

The experimental results indicate that the likelihood of rejection of an offer is an increasing function of the asymmetry of P’s offer (measured by the fractional differences between P and R entailed), but is not very sensitive to the size of the pie (within experimental ranges) or to the nationalities, cultures or sub-cultures from which the players are drawn.

The results suffice to falsify a line of economic reasoning that Rs should accept any offer that leaves them with any slice of the pie, however small, on the ground that the offer is a given fact that can’t be changed (bygones are bygones) and something is better than nothing. Moreover, the relevant evidence indicates that this behaviour cannot be dismissed as being a consequence of irrationality or stupidity.

The structure of the UG is characterised by an inbuilt asymmetry of power: P has more influence on outcomes than R, manifested in the fact that, averaged over experiments as a whole, the Ps leave the room with more pie. Yet for either party to get anything at all, both must approve the proposal. How that greater power is exercised by the Ps is, we think, of normative concern to the Rs. As recorded by St Luke: “For unto whomsoever much is given, of him shall be much required …”

We suggest that the UG results are most readily explicable by the ubiquity of this normative power-responsibility linkage, which points to a disapproving appraisal of the conduct of a P who makes a highly asymmetrical offer. In the context of a power difference, it implies a lack of due consideration for either the interests or the veto power of the Rs. (More extensive development of the point is set out in a forthcoming RPI paper in the new, ‘Past Learnings’ series.)

What then has this got to do with the regulation of immigration? Our answer is that, though differing in many obvious ways, there are some similarities between the structures of the UG and of electoral politics. Politicians themselves make reference to the “retail offers” they make to the public in the form of policy promises (as if they were proprietors of market stalls). The public examine those offers and can respond to them by crosses on ballot papers, whose positioning is affected by feelings of approval or disapproval. If sufficient levels of disapproval are reached, the offerors can be dispossessed of the benefits of office in consequence (just as the Ps are deprived of their shares of the pie in the UG).

Economic research on the economic effects of immigration is very limited and it tends to focus on assessments of aggregated net benefits. This is ill-directed effort, if it takes zero immigration as the benchmark for assessment. Very few, we think, would favour that: even proponents of net-zero policies approve of inflows of people that match the outflows of emigrants. In practice, the salient policy issues are to do with increments in benefits and costs measured from a status quo position. It is these incremental net benefits that correspond to the size of the pie in the UG.

What the UG suggests, though, is that it is not the size of the pie that is the major determinant of the pattern of rejections, but rather asymmetries in the distribution proposed by the offeror. In the case of immigration policy, the relevant public judgements are clearly about the conduct of politicians in determining the offer, but not, we think, principally about the distribution of net benefits between voters and the politicians themselves.

Rather, we suggest that the prime suspect for an asymmetry capable of driving a sense of unfairness or injustice is the division of the net benefits between a member of the electorate and a newly arrived immigrant. The question is: on average, does the new arrival receive a disproportionally large slice of any net benefit consequential on the change of residence? It is seemingly not a question that is systematically asked at a policy level. The UG evidence suggests that it should be.

Share on Twitter
Share on LinkedIn