Proportionality, winner-take-all, and distributive justice

Mark R. Reiff

Research output: Contribution to journalArticlepeer-review


When faced with multiple claims to a particular good, what does distributive justice require? To answer this question, we need a substantive moral theory that will enable us assign relative moral weights to the parties' claims. But this is not all we need. Once we have assessed the moral weight of each party's claim, we still need to decide what method of distribution to employ, for there are two methods open to us. We could take the winner-take-all approach, and award the good to the party with the strongest claim. On the other hand, we could divide the good proportionally, according to the relative strength of each party's claim. Because the choice between these two methods of distribution can have a dramatic impact on the resulting pattern of distribution, the choice presents a question of justice. But this is a question of justice that is often overlooked. As a result, we currently employ the principle of proportionality far less often than justice actually requires. If we focus on the question of distributive method, however, we are not only better able to understand how certain reasons enter into our all-things-considered moral judgments, we are also able to explain some perplexing but common aspects of our moral beliefs: how rights can be said to have peremptory force, yet still be balanced against other important interests; how justice can sometimes require compromise, yet sometimes require victory; and how a moral theory can avoid being too demanding while still being demanding enough. © SAGE Publications Ltd.
Original languageEnglish
Pages (from-to)5-42
Number of pages37
JournalPolitics, Philosophy and Economics
Issue number1
Publication statusPublished - Feb 2009


  • Autonomy
  • Equality
  • Indivisibility
  • Inviolability
  • Rights
  • Weighted lotteries


Dive into the research topics of 'Proportionality, winner-take-all, and distributive justice'. Together they form a unique fingerprint.

Cite this