The middle ground is generally used to describe a place between two or more extremes. Utilitarianism has historically been associated with maximizing expected utility, but there is another decision procedure that could conceivably merit consideration. This is the maximin principle, which was most famously advocated by John Rawls in A Theory of Justice as the principle to use when behind the Veil of Ignorance.
Maximin Utilitarianism is an interesting, if limited idea. It does in some sense capture our moral intuitions about rights very well, as well as the distinction between persons. However, it is perhaps too rigid, and by itself, does not consider the greater good.
Perhaps we can combine this with a regular form of Utilitarianism to create something more balanced.
Maximin Balanced Average Utility = Average(Minimum(x), Average(x))
What does this do? In effect it balances between the minimum utility and the average utility. Thus, imagine a situation where one person must suffer a year of torture in exchange for avoiding specks of dust in 10 billion people’s eyes. Under a regular Utilitarian framework, the sheer number of people affected would justify this trade. However, under Middle Ground Utilitarianism, the year of torture has far greater magnitude in terms of suffering than the average magnitude of the dust in people’s eyes, and so Middle Ground Utilitarianism would reject this trade.
Notice a few salient points. By taking the average rather than the total, we make the comparison between any given individual person’s suffering equal. In some sense this captures the distinction between persons better than just adding up all the suffering. Many small pinpricks over space and time cannot add up to be more relevant than a single moment of intense agony.
Thus, Middle Ground Utilitarianism avoids suggesting that we should hang an innocent man to appease a large mob, among many other similar examples of where Utilitarianism historically has made people uncomfortable with its implications. And all it took was to give equal consideration to maximin and expected utility as decision frameworks.
In other cases, Middle Ground Utilitarianism still supports maximizing the good. For instance, with the trolley problem, the minimum is equally bad for either the one or the five, so the average works out as a tie breaker and supports saving the five at the expense of the one.
In other situations, like population ethics, it works out to be similar to Average Utilitarianism. In this sense, it is indifferent to the number of people, and cares primarily about average happiness. One point of value for Average Utilitarianism is that it handles the issue of Infinite Ethics better than Total Utilitarianism. Middle Ground Utilitarianism should perform similarly.
Thus, I offer yet another alternative implementation of Utilitarianism, one that perhaps better matches our moral intuitions and conceptions of justice, fairness, and rights.