Social Contract Theory and Future Generations
Societies are not static; they are always changing. The principles of justice—the principles for how to organize society in a just manner—must be able to be applied repeatedly to address new injustices. This dynamic aspect of justice has received very little attention. One aim of the project is to explore how Social Contract Theory can handle changes in society—changes in technology, circumstances, and the addition of future generations. A second aim is to explore dynamic decision-making not just in Social Contract Theory but also in population ethics, the part of ethics that deals with cases where different people exist in different potential outcomes. Political policies that concern the future—e.g., policies for global warming—do not just affect the quality of future lives, they also affect who will exist in the future. Earlier work in distributive justice and population ethics has looked at static choices—single choices between different ways to organize society or between different futures. This restriction to static cases is unrealistic. We do not make a single choice of one way of organizing society and then stick to it for all time. Rather, the way we organize society is constantly re-evaluated through elections and other social changes. And we do not make just one choice that determines what the future will be like. The future is chosen through a sequence of choices over time. The project will explore the problems that are raised by this dynamic aspect of justice.
Final report
# Final report
## Purpose
The purpose of the project was to investigate how social contract theories can handle changes in society. For instance, culture and technology can change and new generations come into being. This means that the principles of justice provided by social contract theories need to be applied multiple times. In social contract theory, the correct principles of justice are those principles we would choose behind a veil of ignorance, where we do not know who we are in society. Behind the veil, individuals do not choose based on considerations of justice but so as to best secure their own interests during their lives. Given that the principles may need to be applied multiple times throughout individuals' lives, individuals need to consider the dynamic effects that iterated applications give rise to. This means that individuals need to make use of dynamic decision theory — that is, how one rationally chooses in sequences of choices.
Besides the fact that sequences of choices are relevant, social contract theory must take into account that the choice of a principle of justice can affect which individuals will exist and what welfare future generations will have. If individuals can affect which principle of justice will be applied in society and this in turn affects who will exist in later generations, it seems that the individuals can conclude that they do not belong to those later generations. They can do this because it seems impossible for the individuals to affect whether they themselves will come into existence (that is, affect whether they are ever born). This leads to individuals becoming biased. They know they do not need to care about later generations (since they are focused on their own welfare).
A second focus of the project has been to explore sequential choices for ethical theories and principles in general. Principles that seem reasonable in static, non-sequential cases can lead to outcomes that are worse for everyone involved in sequential cases. These cases are called value pumps, as the application of the principles makes everyone worse off than if the principles were not applied. More generally, the project has also explored principles of rationality and sequential choices to avoid money pumps — that is, cases where the application of a principle leads to paying certain cost for the same outcome as if no principles had been applied but without paying the cost. This is more general because what one prefers can be anything — for example, one's own welfare (as is the case for the individuals behind the veil of ignorance) or what is morally valuable.
## Results
To avoid the problem that individuals behind the veil of ignorance could figure out that they belong to an early generation if they can (causally) affect which principle of justice will be applied in society, I have explored (in joint work with Andreas Mogensen) a variant of social contract theory where individuals behind the veil of ignorance do not (causally) influence which principle is applied in society. Instead, their choice of principle is merely (statistical) evidence that the same principle has been applied in society. For this to be relevant to the individuals, they need to use evidential decision theory, where alternatives are evaluated solely based on their evidential value. (That is, one chooses the alternative one would most like to learn one has chosen.)
Given this variant, we avoid problems with individuals being biased in favour of early generations. Another problem is what probability distribution individuals should have for being the different persons in the different outcomes. The proposal that avoids value pumps is that, in each outcome, individuals have the same probability for all persons in the outcome of being that person in that outcome. This leads, given that individuals value risk according to the expected utility theory, to them choosing the principle of justice that arranges society so that the expected average welfare is as high as possible.
Besides this, I have defended (in joint work with Kacper Kowalczyk) the underlying assumption that a reasonable principle of justice does not give all involved lower expected welfare, in response to influential criticism by Anna Mahtani.
In the second part of the project, I have shown that reasonable ethical theories must take into account the interests of merely possible future individuals. It is not enough to only consider the interests of the future individuals who will actually exist. And I have shown that prioritarianism (a variant of utilitarianism that gives more weight to improvements for those who are worse off) can in some sequential cases lead to all individuals receiving lower welfare.
More generally, I have shown that rational preferences in sequential choices must be in accordance with expected utility theory. Otherwise, the preferences can be exploited in a money pump. This led to the book Money-Pump Arguments. This supports the assumption that the individuals behind the veil of ignorance in social contract theory evaluate principles based on their expected value.
## New questions
Two surprising findings were that the distinction between evidential and causal decision theory is crucial for social contract theory, and also that anthropic reasoning matters (if a principle of justice — based on, for instance, a coin flip — lead to many future generations or to very few generations, is it then more likely that you belong to one of the many generations or one of the few?). These connections between social contract theory and evidential/causal decision theory as well as anthropic reasoning have not been explored before.
## International dimensions
The project has had a strongly international profile. Not just among the participants at the concluding workshop, but also among the co-authors of the articles from the project. I visited the Global Priorities Institute at University of Oxford and wrote the article ‘Future Generations and the Evidential Veil of Ignorance’ with Andreas Mogensen (Oxford). I also visited the Population Wellbeing Initiative in Austin, Texas and wrote the article ‘Ex-Post Average Utilitarianism Can Be Worse for All Affected’ with Dean Spears (UT Austin). Additionally, I collaborated with Kacper Kowalczyk (UCL) which resulted in the article ‘Ex-Ante Pareto and the Opaque-Identity Puzzle’.
## Dissemination
My results from the project have been disseminated at conferences at the University of Texas at Austin, Institute for Futures Studies, University of Konstanz, University of Nebraska-Lincoln, University of Stirling, Texas A&M, Erasmus University Rotterdam, and University of Oxford.
## Purpose
The purpose of the project was to investigate how social contract theories can handle changes in society. For instance, culture and technology can change and new generations come into being. This means that the principles of justice provided by social contract theories need to be applied multiple times. In social contract theory, the correct principles of justice are those principles we would choose behind a veil of ignorance, where we do not know who we are in society. Behind the veil, individuals do not choose based on considerations of justice but so as to best secure their own interests during their lives. Given that the principles may need to be applied multiple times throughout individuals' lives, individuals need to consider the dynamic effects that iterated applications give rise to. This means that individuals need to make use of dynamic decision theory — that is, how one rationally chooses in sequences of choices.
Besides the fact that sequences of choices are relevant, social contract theory must take into account that the choice of a principle of justice can affect which individuals will exist and what welfare future generations will have. If individuals can affect which principle of justice will be applied in society and this in turn affects who will exist in later generations, it seems that the individuals can conclude that they do not belong to those later generations. They can do this because it seems impossible for the individuals to affect whether they themselves will come into existence (that is, affect whether they are ever born). This leads to individuals becoming biased. They know they do not need to care about later generations (since they are focused on their own welfare).
A second focus of the project has been to explore sequential choices for ethical theories and principles in general. Principles that seem reasonable in static, non-sequential cases can lead to outcomes that are worse for everyone involved in sequential cases. These cases are called value pumps, as the application of the principles makes everyone worse off than if the principles were not applied. More generally, the project has also explored principles of rationality and sequential choices to avoid money pumps — that is, cases where the application of a principle leads to paying certain cost for the same outcome as if no principles had been applied but without paying the cost. This is more general because what one prefers can be anything — for example, one's own welfare (as is the case for the individuals behind the veil of ignorance) or what is morally valuable.
## Results
To avoid the problem that individuals behind the veil of ignorance could figure out that they belong to an early generation if they can (causally) affect which principle of justice will be applied in society, I have explored (in joint work with Andreas Mogensen) a variant of social contract theory where individuals behind the veil of ignorance do not (causally) influence which principle is applied in society. Instead, their choice of principle is merely (statistical) evidence that the same principle has been applied in society. For this to be relevant to the individuals, they need to use evidential decision theory, where alternatives are evaluated solely based on their evidential value. (That is, one chooses the alternative one would most like to learn one has chosen.)
Given this variant, we avoid problems with individuals being biased in favour of early generations. Another problem is what probability distribution individuals should have for being the different persons in the different outcomes. The proposal that avoids value pumps is that, in each outcome, individuals have the same probability for all persons in the outcome of being that person in that outcome. This leads, given that individuals value risk according to the expected utility theory, to them choosing the principle of justice that arranges society so that the expected average welfare is as high as possible.
Besides this, I have defended (in joint work with Kacper Kowalczyk) the underlying assumption that a reasonable principle of justice does not give all involved lower expected welfare, in response to influential criticism by Anna Mahtani.
In the second part of the project, I have shown that reasonable ethical theories must take into account the interests of merely possible future individuals. It is not enough to only consider the interests of the future individuals who will actually exist. And I have shown that prioritarianism (a variant of utilitarianism that gives more weight to improvements for those who are worse off) can in some sequential cases lead to all individuals receiving lower welfare.
More generally, I have shown that rational preferences in sequential choices must be in accordance with expected utility theory. Otherwise, the preferences can be exploited in a money pump. This led to the book Money-Pump Arguments. This supports the assumption that the individuals behind the veil of ignorance in social contract theory evaluate principles based on their expected value.
## New questions
Two surprising findings were that the distinction between evidential and causal decision theory is crucial for social contract theory, and also that anthropic reasoning matters (if a principle of justice — based on, for instance, a coin flip — lead to many future generations or to very few generations, is it then more likely that you belong to one of the many generations or one of the few?). These connections between social contract theory and evidential/causal decision theory as well as anthropic reasoning have not been explored before.
## International dimensions
The project has had a strongly international profile. Not just among the participants at the concluding workshop, but also among the co-authors of the articles from the project. I visited the Global Priorities Institute at University of Oxford and wrote the article ‘Future Generations and the Evidential Veil of Ignorance’ with Andreas Mogensen (Oxford). I also visited the Population Wellbeing Initiative in Austin, Texas and wrote the article ‘Ex-Post Average Utilitarianism Can Be Worse for All Affected’ with Dean Spears (UT Austin). Additionally, I collaborated with Kacper Kowalczyk (UCL) which resulted in the article ‘Ex-Ante Pareto and the Opaque-Identity Puzzle’.
## Dissemination
My results from the project have been disseminated at conferences at the University of Texas at Austin, Institute for Futures Studies, University of Konstanz, University of Nebraska-Lincoln, University of Stirling, Texas A&M, Erasmus University Rotterdam, and University of Oxford.