Redistributive policy

Universal benefits cost less than means-tested benefits – People’s Policy Project




In 1998, Walter Korpi and Joakim Palme published “The Paradox of Redistribution and Equality Strategies: Welfare State Institutions, Inequality, and Poverty in Western Countries.” The paper aimed to tackle the argument that targeting welfare programs to people with low market incomes reduces inequality and poverty more than not targeting them in this way.

In their response, Korpi and Palme accept the basic calculations of the “targeting effectiveness” claim, but argue that they ignore the political effects of such targeting on support for increased size of the social budget. If (1) large, untargeted welfare states reduce inequality and poverty more than small, targeted welfare states, and (2) targeting makes welfare states small, then (3) targeting is actually worse for inequality and poverty than non-targeting.

This paper started a fire in welfare state discourse that still burns today. A large number of secondary papers have been written on the subject trying to prove or disprove its thesis with various cross-country regressions. And the basic argument has become boilerplate rhetoric for social democratic politics around the world.

But Korpi and Palme’s paper makes a key mistake that few have taken note of. In fact, once this error is understood, the whole argument of the article and the discourse it spawned become moot.

Korpi and Palme accept the idea that, for a given social budget, targeted programs reduce inequality and poverty more than non-targeted programs. This claim is not true, and the seemingly irrefutable math behind it is based on counting games.

The Greatest Trick in Welfare History

My favorite way to dissect the “targeting effectiveness” argument is to imagine two people sitting at a table, one a proponent of untargeted programs (i.e. universalism) and the another an advocate for targeted programs (ie means testing). A sheet of paper is given to each of them with the following challenge:

  1. Design a tax policy that generates exactly $100 billion in revenue. No additional tax is allowed.
  2. Design a cash welfare program for children that costs exactly $100 billion. No additional expenses allowed.
  3. The combination of fiscal and social policies that reduces poverty and inequality the most wins.

After receiving the challenge, the Resource Tester and Universalist begin working on their $100 billion tax policies.

The two examine the criteria for victory – reducing poverty and inequality the most – and independently design the same tax policy that optimally levies $100 billion from the wealthiest people in society. After a few minutes, each reveals their tax policy to the other and laughs that they found the same thing.

Then the resource tester and universalist begin work on their $100 billion child welfare package.

The Universalist, being a Universalist, naturally decides to take the $100 billion, divide it by the number of children in society, and then pay out the resulting dollar amount to each child. According to the calculations of the Universalist, this equates to $3,000 per child.

The resource tester, being a resource tester, takes a different approach. Rather than providing $3,000 to each child, the resource tester decides to provide $6,000 per child, except that children living in families earning more than $50,000 a year will receive 5 cents less than the 6 benefit. $000 for every dollar they earn above $50,000.

After a few minutes, each reveals his social policy to the other. The Resource Controller smiles and triumphantly declares that his social policy provides twice as many benefits to poor and low-income children as the Universalist’s social policy. Since the Universalist and the Resource Controller have the same tax policy, this difference in social policy makes the Resource Controller the winner of the challenge.

After seeing the Resource Tester’s Wellness Policy, the Universalist immediately begins revising both of his policies. It adds a second line to its tax policy that says “5% surtax on income over $50,000 for tax units with children, up to $6,000 tax per child.” It changes its welfare policy to increase the benefit from $3,000 per child to $6,000 per child.

Once the Universalist reveals these changes, a heated discussion ensues:

Means-Tester (MT): The rules clearly state that taxes and expenses cannot exceed $100 billion. Under your revised plan, you collect $200 billion in tax revenue and spend $200 billion on the $6,000 Universal Child Benefit. You are disqualified.

Universalist (UN): In your spending plan, parents who earn more than $50,000 in income lose 5 cents in benefits for every dollar they earn over $50,000. This “phasing out” is exactly the same as my “tax”. Either both violate the rules that taxes and expenditures cannot exceed $100 billion, or neither.

MT: No, phasing out is not a tax! It is a reduction in expenses. I taxed and spent $100 billion to provide a $6,000 means-tested benefit. You taxed and spent $200 billion to provide a $6,000 universal benefit.

UN: Nope. We both taxed and spent $200 billion to provide a $6,000 universal benefit. It’s just that your 5% tax is applied at the time of the benefit payment, while my 5% tax is applied at the time of the paycheck payment. Other than this administrative difference, our policies are exactly the same. In fact, come to think of it, applying income tax when paychecks are paid is administratively easier than applying income tax when benefits are paid. This means that my policy requires less bureaucratic expense than yours. I change my social policy. My lower administrative costs allow me to increase my benefits to $6,200 per child. I am now giving $200 more to poor children than you and I am the winner of the challenge.

The Resource Controller and the Universalist fail to settle this dispute, and so both of their policies go to trial. Unfortunately, when selecting a judge, the challenge organizers looked for someone with extensive experience handling tax and expense microsimulations. The judge is able to simulate both sets of policies and conclude that the calculations verify both: the means-tester can really finance a means-tested benefit of $6,000 with his tax on the rich and the universalist can really fund a $6,200 universal benefit with its tax on the wealthy and a 5% tax on families with incomes over $50,000. From microsimulations alone, the universalist has won.

But the resource tester insists the Universalist’s policy violates the rule that taxes and spending cannot exceed $100 billion. The resource evaluator no longer argues that his policy is optimal in any welfare-maximizing sense of the term, but rather argues that it satisfies the semantic requirements of the challenge rules. The judge does not know how to settle this semantic dispute.

Eventually, it was decided that the only way to settle it was to determine how the national accounts would categorize the amounts of money in the two policy packages. After a few hours of browsing national accounts guides, a clear answer emerges. Under existing national accounting rules, the phasing out of the resource tester is counted as a reduction in spending while the distributively identical and administratively higher tax of the universalist is counted as an increase in taxes. The Universalist is disqualified and the Means Tester is declared the winner.

Reject the premise

Korpi and Palme’s mistake was not realizing that “targeting” is just taxing under another name. Resource testers have not figured out how best to spend a fixed amount of tax revenue. Rather, in these debates, they use national accounting rules to allow themselves to tax more in order to spend more while preventing Universalists from doing the same. Simply put: the whole idea of ​​“targeting efficiency” and the simple math behind it is just a game of bookkeeping.

Korpi and Palme’s argument that targeting policy results in smaller welfare states that are, ultimately, worse for the poor is very elegant, and they put together impressive international statistics in its favour. But this is also not necessary at all. The assumption that, relative to universal programs, targeted programs result in more aid to the poor for the same amount of expenditure is quite wrong and must be rejected.

Amusingly, once the mechanics of “targeting efficiency” are truly understood, it is the means testers, not the universalists, who ultimately have to rely on Korpi and Palme-type arguments about politics. if they hope to argue any type of case for their technically. inferior and more costly political designs. Specifically, resource testers must ultimately assume that, despite their distributive similarities, the public reacts differently to a phase-out than to a tax, and that the policy possibilities opened up by this public confusion allow for much more taxation (through phase-outs progressive) and spending than universalist alternatives that it overcomes targeting inefficiencies.

So far, no one has had to make such an argument, because almost no one who argues on these topics really understands that “targeting effectiveness” is an accounting gimmick. But maybe one day this understanding will be more widespread and we will get a landmark article on the paradoxical way in which implementing taxes as phase-outs, despite administrative waste and a drain on the poor people’s participation in programs, however, results in more benefits for the poor than more effective one-size-fits-all designs.