A Revival of the Marshallian Approach to Welfare

With respect to the theory of welfare, Robertson assumed the overtly Marshallian form of utility theory to defend Marshall’s approach in the article ‘On Sticking to One’s Last’ (Robertson 1949 [1954a]). In Robertson’s opinion, the suppression of utility from the discussions on welfare theory, as recommended by Robbins and Hicks in the 1930s, had hampered economists’ ability to provide useful advice on substantive questions affecting society. This regrettable situation resulted from economists being deprived of a simple and powerful analytical tool in the evaluation of how legislation or other forms of intervention would impact people’s material well-being. Without the concept of utility, economists had embraced distinct and fancy criteria of social improvement which emboldened some to conceive unfeasible schemes of delivering Heaven on Earth. Instead, said Robertson, economists should be humble and aim just at presenting a fair case to society, wishing their arguments were properly understood and knowing that a final decision on any matter regarding welfare would most likely take other factors into consideration (ibid.: 60-65).

In ‘Utility and All That’ (Robertson 1951 [1954b]), Robertson frontally attacked the so-called Hicks-Kaldor compensation principle of welfare optimality. According to this criterion, anchored in the Paretian tradition of thought, a change in income distribution could only be said to have increased welfare if those individuals whose situation had improved were able to monetarily compensate the ones whose condition had worsened and, notwithstanding this, still be left with some gain. The main flaw of this line of reasoning, argued Robertson, was the fact that any evaluation of what actually happens depends on what might happen, but most likely will not happen anyway. ‘Potential’ does not mean ‘actual’ and given this, the new income distribution under scrutiny would end up being assessed by an ethical standard foreign to economics. This limitation, however, does not affect the analysis based on the cardinal utility assumption, which allows the economist to go further away and reaffirm the Pigouvian conclusion that a less unequal income distribution increases overall welfare (ibid.: 30-41). Formal elegance, advised Robertson, was no substitute for common sense when it comes to social issues: ‘But when we reach these regions of ultimate decision, I find it difficult to believe that mathematics has much to say that is not said more briefly in that aspiration of the book of Common Prayer - that we may have a right judgement in all things’ (ibid.: 41).

As Robertson reaffirmed the validity of employing the notion of cardinal utility in welfare economics, he pointed out as well a possible inconsistency in Pigou’s approach to the subject. In Wealth and Welfare, it is said that the growth in the national dividend brings forward an increase in welfare whenever it does not reduce the share of the national dividend accruing to the poor. Robertson, however, warned that if distribution is affected by the overall performance of the economy, it may happen that a more equal distribution in income could be accompanied by a contraction in the national dividend, leaving the final effect on welfare undetermined (ibid.: 30-37; Boianovsky 2014).

Finally, in the article ‘Utility and All What?’ (Robertson 1954), Robertson kept pointing out that the neglect of cardinal utility had critically weakened welfare economics. This time he was critical of Robbins. Economists, said Robertson, can and should provide valuable advice about the effects of economic changes on human happiness, particularly on those policies which impact income distribution. To omit oneself from such a noble duty, as suggested by Robbins, was actually a defeatist position which lowered the standing of economics within society. Additionally, this passive attitude left the field clear for others to offer advice on crucial matters involving the application of scarce resources among alternative uses, a problem which Robbins himself had taught was at the core of economic science (ibid.).