The small-mindedness of means-testing
The hot topic in British politics last week was the government’s decision to scrap the winter fuel payment. People over the age of 65 used to be able to claim a lump sum of between £200 and £300 pounds each winter. Desperately scrabbling around for cash, the government has changed the policy so that now only elderly people who are already receiving state financial help are eligible for the payment. This is a classic example of “means-testing”: making state benefits only available to those who do not have the means to pay for things themselves.
Means-testing tends to be popular because it seems to make a lot of sense. Why waste money providing benefits to millionaires? At the most general level, a state with any egalitarian ambitions must treat the rich and poor differently.
Nonetheless, means-testing is generally small-minded and regrettable.
One reason is to do with political feasibility. Universally provided benefits have lots of people to fight for them. Means-tested benefits, by their nature, only benefit a small group. When governments need cash fast, these benefits are easy targets because few people know about them and are motivated to fight for them. One could also add a more principled gloss to this feasibility argument by claiming that universal benefits express social solidarity.
What gets me though is that while means-testing is justified as being more efficient, it’s generally completely the opposite. Universal benefits look inefficient from an accountant’s perspective: they involve large transfers of money. But from an economist’s perspective, large transfers don’t cost anything at all – they don’t involve using resources, just moving them around. What actually uses resources are the administrative frictions of making the transfers. Universal benefits create few frictions: the applications are easy to fill out, the work for civil servants is minimal. Means-tested programmes generate lots of friction: the application forms are confusing and time-consuming, and sorting it all out makes more work for civil servants.
Of course, if we care about poverty or inequality the state will have to put in the work of figuring out people’s needs and resources at some point. But the state already does this in the tax system. Doing it again in the administration of state benefits is inefficient. Rather than means-testing benefits individually, you can get functionally more or less the same result from a system that provides benefits universally and levies higher and more progressive taxes. The universalist system will look inefficient, because the state will tax the rich with one hand and then pay them back a lot of that money in the form of universal benefits on the other hand. But because these transfers involve less friction, the whole system will waste fewer resources on bureaucracy than a system based on means-testing.
To be clear, I’m not saying the original winter fuel policy was a brilliant policy. Arguably, it was already too targeted: don’t poor people under 65 get cold in the winter as well? If you take this line of thought seriously enough, you end up with universal basic income.
Means-testing is a manifestation of what we might call the mindset of localised fairness: worrying about the fairness of particular policies/institutions/practices without considering how they fit into the whole social system. Is there a similar tendency in contemporary liberal egalitarian political theory? We are asked to worry about justice in the distribution of a multitude of particular goods (leisure time, clean air, healthcare), or within particular institutions (workplaces). Of course, there are lots of arguments about why particular kinds of fairness should be considered separately, but I wonder if theorists fall into a mindset of localised fairness as they search for new nails with which to whack with the hammer of distributive justice.
