What do you think?
Rate this book
California Series on Social Choice and Political Economy
Choosing Justice: An Experimental Approach to Ethical Theory (California Series on Social Choice and Political Economy)
This book presents an entirely new answer to the question: “What is fair?” In their radical approach to ethics, Frohlich and Oppenheimer argue that much of the empirical methodology of the natural sciences should be applied to the ethical questions of fairness and justice.
276 pages, Paperback
First published June 24, 1992
15 people want to read
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 of 1 review
January 19, 2022
Can we put justice under the microscope? Frohlich and Oppenheimer immodestly claim to have found a robust method of turning ethics into an experimental, empirical science that produces more trustworthy (indeed, "truer") ethical conclusions than the traditional methods of armchair theorizing, such as intuitionism and introspective analysis. Their revolutionary ambitions and substantive ethical conclusions can be questioned, but their fresh methodological approach to ethical reasoning is a fountainhead of insights and promising experimental pathways.
Frohlich & Oppenheimer start from the Rawlsian-Harsanyian premise (which is also found, for example, in the Kantian and utilitarian traditions) that rules of justice require the approximation of conditions of "impartial reasoning" where everybody's interests are given equal weight and where rules of social interaction are decided from behind a "veil of ignorance" where nobody knows how well off or badly off they will be in the society. So far so Rawlsian. The methodological innovation of Frohlich & Oppenheimer lies in the additional controversial claim that such conditions can be approximated (best) in a laboratory setting. They correctly point out the fact that Rawls, Harsanyi, and other contractarian theorists, although they agree on the methodological premise, cannot agree on the precise principles of justice that are supposed to emerge from behind the veil of ignorance. There is no agreed upon substantive outcome that emerges from the "original position." From this F&O conclude that it is important to observe empirically what people WILL, in fact, choose. And since we cannot usually recreate or replicate the veil of ignorance in the political order in which we live, they continue, the best available approximation lies in carefully designed laboratory conditions where small groups of people are allowed to freely deliberate and make a collective choice about rules of income distribution from behind a veil that obscures each agent's expected social payoffs and work capacity. Such a scenario is supposed to replicate some (but not all) features of the Rawlsian-Harsanyian veil. The authors therefore conclude that lab experimentation constitutes a robust method of testing out the popularity (and therefore, ex hypothesi, desirability) of different theories of (impartial) justice. Ethics, therefore, can be subjected to empirical verification or, at least, in proper Popperian fashion, falsification: those theories that are repeatedly rejected by lab experimental conditions should be rejected, or at least doubted, as good theories of justice.
The empirical data gathered from their experiments is very interesting. The authors report that, based on over seventy experiments in three countries, people tend to converge around similar principles of (distributive) justice. This is surely the most interesting result of their work. It is very noteworthy that Rawls's difference principle, which maximizes the wellbeing of the least well-off group, fares very poorly in empirical settings. Likewise, Harsanyi's utility-maximizing principle, which focuses on maximizing the average social payoff, rarely emerges as the preferred principle of justice from behind the (lab-simulated) veil of ignorance. Instead, most groups gravitate around average utility maximization constrained by a guaranteed income floor below which no person can fall. In other words, income utilitarianism with a "floor constraint." Although the authors never use the word, this suggests something like a guaranteed "basic income" floor as the ideal constraint on an inequality-engendering social order where the income distribution is otherwise free to fluctuate in a way that is sensitive to people's ambitions, skills, and preferences.
All of this is very interesting. Does it hold up? I will not attempt a thorough critique here but let me point out a few ways in which the laboratory settings under which the experiments ran were full of shortcomings. First, 1) the uncertainty achieved in the lab experiments was not as deep as the uncertainty demanded by Rawls in his thought experiments; 2) the lab experiments were conducted in groups of no more than five people, which could have contributed to "excessive" levels of consensus-building, trust, and sympathy compared to decision-making in a diverse, populous society; and 3) perhaps most importantly, the lab experiments were primarily concerned with optimal methods of income redistribution (from a given pie) and only secondarily with economic production (the baking of the pie), based on work incentives, capital accumulation, social innovation, etc. Indeed, economic production can be modelled, if at all, only very reductively and imperfectly in a laboratory setting. A laboratory can never behave like a complex economy.
There are also other problems and limitations of laboratory experimentation, some of which are anticipated and discussed (although not fully resolved) by the authors themselves. It seems to me that the authors' noble efforts at devising robust experimental laboratory conditions suitable for measuring moral "truth" ultimately fall short of robustly approximating the conditions of impartial reasoning, but there is much to be learned from their efforts. The substantive results are suggestive that some "mixed" or "hybrid" principle of distributive justice can garner wide appeal from a multitude of perspectives. This leads some credence to efforts to devise concrete institutional reforms based on the "floor constraint" principle, such as UBI, NIT, and minimum income. On a formal level, beyond the substantive conclusions, lab experiments can be useful and interesting across several margins. They can be used to simulate contractarian and impartial ethical-political reasoning, to model deliberative mini-publics, to study human social psychology, and to observe how different experimental parameters affect (or fail to affect) people's ethical views and social choices. The continuing usefulness of lab experiments as tools in the pluralistic tool kit of moral philosophers (especially those working in the contractarian tradition), although unlikely to outperform centuries of "arm chair" theorizing in philosophy, is unquestionable.
Frohlich & Oppenheimer start from the Rawlsian-Harsanyian premise (which is also found, for example, in the Kantian and utilitarian traditions) that rules of justice require the approximation of conditions of "impartial reasoning" where everybody's interests are given equal weight and where rules of social interaction are decided from behind a "veil of ignorance" where nobody knows how well off or badly off they will be in the society. So far so Rawlsian. The methodological innovation of Frohlich & Oppenheimer lies in the additional controversial claim that such conditions can be approximated (best) in a laboratory setting. They correctly point out the fact that Rawls, Harsanyi, and other contractarian theorists, although they agree on the methodological premise, cannot agree on the precise principles of justice that are supposed to emerge from behind the veil of ignorance. There is no agreed upon substantive outcome that emerges from the "original position." From this F&O conclude that it is important to observe empirically what people WILL, in fact, choose. And since we cannot usually recreate or replicate the veil of ignorance in the political order in which we live, they continue, the best available approximation lies in carefully designed laboratory conditions where small groups of people are allowed to freely deliberate and make a collective choice about rules of income distribution from behind a veil that obscures each agent's expected social payoffs and work capacity. Such a scenario is supposed to replicate some (but not all) features of the Rawlsian-Harsanyian veil. The authors therefore conclude that lab experimentation constitutes a robust method of testing out the popularity (and therefore, ex hypothesi, desirability) of different theories of (impartial) justice. Ethics, therefore, can be subjected to empirical verification or, at least, in proper Popperian fashion, falsification: those theories that are repeatedly rejected by lab experimental conditions should be rejected, or at least doubted, as good theories of justice.
The empirical data gathered from their experiments is very interesting. The authors report that, based on over seventy experiments in three countries, people tend to converge around similar principles of (distributive) justice. This is surely the most interesting result of their work. It is very noteworthy that Rawls's difference principle, which maximizes the wellbeing of the least well-off group, fares very poorly in empirical settings. Likewise, Harsanyi's utility-maximizing principle, which focuses on maximizing the average social payoff, rarely emerges as the preferred principle of justice from behind the (lab-simulated) veil of ignorance. Instead, most groups gravitate around average utility maximization constrained by a guaranteed income floor below which no person can fall. In other words, income utilitarianism with a "floor constraint." Although the authors never use the word, this suggests something like a guaranteed "basic income" floor as the ideal constraint on an inequality-engendering social order where the income distribution is otherwise free to fluctuate in a way that is sensitive to people's ambitions, skills, and preferences.
All of this is very interesting. Does it hold up? I will not attempt a thorough critique here but let me point out a few ways in which the laboratory settings under which the experiments ran were full of shortcomings. First, 1) the uncertainty achieved in the lab experiments was not as deep as the uncertainty demanded by Rawls in his thought experiments; 2) the lab experiments were conducted in groups of no more than five people, which could have contributed to "excessive" levels of consensus-building, trust, and sympathy compared to decision-making in a diverse, populous society; and 3) perhaps most importantly, the lab experiments were primarily concerned with optimal methods of income redistribution (from a given pie) and only secondarily with economic production (the baking of the pie), based on work incentives, capital accumulation, social innovation, etc. Indeed, economic production can be modelled, if at all, only very reductively and imperfectly in a laboratory setting. A laboratory can never behave like a complex economy.
There are also other problems and limitations of laboratory experimentation, some of which are anticipated and discussed (although not fully resolved) by the authors themselves. It seems to me that the authors' noble efforts at devising robust experimental laboratory conditions suitable for measuring moral "truth" ultimately fall short of robustly approximating the conditions of impartial reasoning, but there is much to be learned from their efforts. The substantive results are suggestive that some "mixed" or "hybrid" principle of distributive justice can garner wide appeal from a multitude of perspectives. This leads some credence to efforts to devise concrete institutional reforms based on the "floor constraint" principle, such as UBI, NIT, and minimum income. On a formal level, beyond the substantive conclusions, lab experiments can be useful and interesting across several margins. They can be used to simulate contractarian and impartial ethical-political reasoning, to model deliberative mini-publics, to study human social psychology, and to observe how different experimental parameters affect (or fail to affect) people's ethical views and social choices. The continuing usefulness of lab experiments as tools in the pluralistic tool kit of moral philosophers (especially those working in the contractarian tradition), although unlikely to outperform centuries of "arm chair" theorizing in philosophy, is unquestionable.
Displaying 1 of 1 review