Who is Welfare For?
Continuing from my last piece, let’s talk about who the focus of welfare should be. We determined that we should be focusing on redistributing to those with the least income. As mentioned earlier, you can only make money off of two things under capitalism alone: labor and owning capital. Ownership of capital is highly concentrated (don’t worry, I’ll fix that in later articles). In fact, the median net worth of the bottom 20% is negative (meaning they have more debt than wealth), while the median net worth of the next 20% is only $10k. This won’t produce enough income to live off of, so the only real option for income for them is labor. Of course, no one working should be in poverty-but we have tools to address that, like raising the minimum wage, unions, and worker ownership. But, as Ryan Cooper notes, “Over three-quarters of poor people are either children, students, disabled, or elderly — that is, people who largely cannot or should not work, which is about half the population.” Here’s a full breakdown of non-workers from this excellent piece.
The market does not provide ANY of these groups with income. And certainly, any sane person would agree that at least the old, young, and seriously disabled shouldn’t have to work. So, what do we do about these non-workers? Simply, the government must step in and give them money. To summarize with a quote from the great Matt Bruenig:
“Due to the fact that half of the population is locked out of market income, there simply is no path to low social inequality that does not rely heavily on the welfare state, aka taxes and transfers.”
From the last post, we already determined that most welfare should be cash. Now we know it should focus on non-workers. Two questions remain. First, should it be universal (given to everyone) or means-tested (given only to the poor)? An example of a universal program is a universal child allowance as proposed by Bruenig, which would give $300 per month per child to everyone in the US. Welfare in the US is generally means-tested, like Medicaid for example. Folks will often argue that this gets money where it’s needed most, while universal programs “waste” money by giving it to the middle-class and rich. But here is a surprising fact: according to studies, the more universal a welfare state is, the MORE it redistributes. This is known as the paradox of redistribution. In other words, targeting welfare leaves the poor worse off. As our goal is to at least end poverty, we can’t have that!
This counter-intuitive conclusion exists for several reasons, and I’ll be drawing from the fantastic James Medlock’s Twitter thread to list them. First, means-testing involves bureaucratic burden to figure out who deserves what. This is a big deal in less developed countries, where such data is harder to come by. But the difference in developed countries is worth noting. For example, in the US, we can compare the universal Old Age, Survivors, and Disibility Insurance (OASDI) to the means-tested Supplemental Security Income (SSI). The latter has about 10 times the fraud rate and administrative costs of the former. Plus means-testing always involves a burden on poor folks, which makes them less likely to use the programs. This is similarly true with tax credits and deductions.
Then you have the fact that means-tested welfare disincentives earning more income compared to a universal system. Say, for example, that some welfare program phases out 40 cents for every extra dollar you make. This is effectively a 40% tax on you, because what matters is not just your market income, but rather your market income + your welfare benefits. It gets worse when you have several interlocking programs, and can lead to dramatic welfare traps. For example:
There’s also an obvious but underrated benefit of universal programs: they are seen as a right, not something you use when you are desperate. No one wants to be seen that way, and lots of welfare has stigma. But getting your Social Security check does not carry that same stigma. We care about this not just because folks should be able to collect welfare and still have dignity and respect, but also because stigma lowers take-up rates for folks who need benefits.
Next is a political practicality, best explained by this meme (no seriously trust me please).
Means-tested welfare pits the middle class, whose taxes fund welfare, against the poor. But a universal, social democratic welfare state aligns the interests of the poor and middle-class, because it benefits them all. Redistribution with a progressive tax system gives more to the middle-class, after all! And the average middle-class person holds more political power than the average poor person, because they simply have a higher financial ability to meaningfully participate in democracy.
This is a good time to note that welfare isn’t just good for redistributing from those well-off right now to those who are not. It’s also good for temporal redistribution, or income smoothing: that is, taking money from you at one point in your life, and giving it back to you at another. This is the basis of Social Security, which is an incredibly successful and beloved program. Studies have shown that people do not naturally smooth their incomes, so welfare is needed the fill the gap. In other words, welfare can also be thought of as social insurance. We pay into unemployment insurance in case we become unemployed. We pay into disability programs because we may become disabled. And so on. This income smoothing brings a lot of folks out of poverty, which is our main goal here. As a prime example, Social Security brings over 21 million Americans a year over the poverty line. But because income smoothing benefits the middle class too, they will support these programs, allowing us to do more to fight poverty.
So, we have come to the counter-intuitive conclusion that trying to target welfare programs at the poor actually hurts them. Now that we know everyone should get welfare, not just the poor, another debate is whether those benefits should be flat (like the aforementioned child allowance) or earnings-related (like Social Security). The answer here varies, but earnings-related benefits undeniably have a crucial role to play. I’ll quote from the original paper on the paradox of redistribution, page 21, with a comparison to corporatist welfare models (where pensions come from your occupation): “Contrary to many scholars’ expectations, earnings-related benefits appear to he a condition for, rather than a hindrance to, the reduction of inequality. Because of their low ceilings for earnings replacement, targeted programs and basic security programs stimulate program exit among she middle class and increase the demand for private insurance. From the point of view of equality, the problem with the corporatist model is not that benefits are earnings oriented. The main difference between the corporatist and the encompassing models is that by organizing the economically active citizens into occupationally segmented social insurance programs, the corporatist model highlights socioeonomic distinctions among different categories of citizens and creates divergent interests among these categories. In contrast, encompassing institutions pool the risks and resources of all citizens and create converging definitions of interest.”
Put another way, earnings-related social insurance is better than the alternative, which would be a very unequal private system. This is why Nordic countries put a lot of effort into crowding private insurance out. “Middle-class families will find that most flat-rate benefits aren’t high enough to meet their standards, so they will turn to the market for additional benefits. Disparities grow between those who have the resources to purchase private pensions, for example, and those who do not. Flat-rate universal benefits merely “establish a base” upon which the market stratification continues to persist, with all its inherent inequalities.”
How much earnings-replacement is required for middle-class buy-in isn’t a question with a defined answer. Social Security, for example, uses some pretty random numbers to determine how much of your earnings are replaced of your working years (90% up to a certain amount, 32% beyond that until a 2nd amount, and 15% beyond that until the cap).
In conclusion: welfare should mainly be for non-workers, not just the poor, and it should have earnings-related components.
