The Case for Social Insurance

Jacob Keegan
7 min readApr 27, 2020


Hey all! Let’s keep talking about welfare. Much of welfare is social insurance: old age insurance (Social Security), unemployment insurance, disability insurance, and health insurance. In my ideal system, this would make up much of total government spending. So it’s important to lay out when social insurance is justified over private insurance, when it isn’t, and how best to distribute it.

Insurance as defined by Google is “a practice or arrangement by which a company or government agency provides a guarantee of compensation for specified loss, damage, illness, or death”. Obviously, there can be private provision of such such things, like flood insurance or auto insurance. But the case for markets is clear there: the risk is obvious and avoidable, high-risk people should be paying more, and it isn’t for everyone. But this isn’t true for all insurance. Generally, justifications for social insurance (that is, government-provided insurance) fall into 4 categories.

  1. Moral arguments. The thing about markets is they always leave people out. This is fine and dandy for speakers or Coke or whatever. But without guaranteed social old age and disability insurance, there would be old and disabled people with no money. While many folks think able-bodied adults should work to get an income, this doesn’t fit the old and disabled. Social Security in the US already lifts 14,810,000 elderly people out of poverty¹, and this could be improved by adding a minimum benefit.

A similar logic applies to health insurance: you shouldn’t be put into bankruptcy just because you got cancer. This is, generally speaking, an event totally out of your control. This idea is even somewhat in our laws: you can no longer be charged higher for health insurance because you have a pre-existing condition. But this means private health insurers can’t price risk, which is the only thing they do well².

It’s also true that income security, which social insurance gives, is directly correlated with economic freedom¹³. Often, social insurance can be a replacement for otherwise constricting regulations. You could try to enforce strict standards of when workers are allowed to be fired… or you can just give them unemployment, for example.

2. Economic efficiency arguments. There are things in life, like disability, that everyone is at a reasonable risk for. So, everyone should have insurance for it. And generally speaking, the bigger the risk pool, the better. This is why Obamacare had compulsive insurance: to make sure low-risk people were in the risk pool, lowering costs for everyone else.

We also must remember that lacking some kinds of insurance places a real constraint on people finding good jobs. The labor market works best if people have high job mobility, and can take the job that pays the most. But there are, to use an economic term, search costs. It takes time to find a job that’s a good fit. But if losing your job means losing health insurance and having no income, you will be more hesitant to push for better pay or find the best job. In fact, making unemployment insurance more generous would increase productivity for this very reason³. A similar argument exists for entrepreneurship: if you want people to take more risks, they need to be protected from risks like not being able to pay for healthcare if their ideas/business fails. With guaranteed health insurance, “Aspiring entrepreneurs would no longer have to worry about getting their health insurance from their spouse, or buying a costly plan on their own in the private market; it would just be there, in the background, providing a safety net that makes the prospect of starting a business less frightening.”

Some folks argue that people should save instead of relying on social insurance for some of these things. But savings for this isn’t efficient over time, for several reasons. First, we can imagine a situation where everyone saved enough for disability, unemployment, or retirement. This would mean saving enough to cover X years of not working, where X is the highest number of years you could be out of work. This, of course, would be an absurdly high amount. And it would be unnecessary: most people would be out of work far less than X years. They would over-save. But if everyone contributed enough to cover anyone who needed it, there would not be over-saving.

Second, some economists have claimed that consumers naturally smooth income over their lives: this is known as the permanent income hypothesis. But this does not fit the data. Since people do not smooth their income well naturally, the welfare state steps in to do it.

Third, young people cannot reasonably save for these outcomes, because they just don’t have enough time to. What if, for example, someone gets out of college, has a job for 2 years, then loses it because the company downsizes?

Lastly, for more thorough results on why savings isn’t optimal even in perfect markets, see this paper by Peter Diamond, or this much more accessible paper.

3. Redistribution. Good welfare, of course, redistributes income from rich to poor. But it also redistributes across age. Just like a dollar is worth more to a poor person than to a rich person, a dollar is worth more to you when you are making the least income in your life than when you are making the most. So, it’s good if you just “redistribute” to yourself. Or rather, you pay into a system that redistributes across age. And, relevant to our discussion, social insurance does this better than private insurance, as long as you have a healthy dose of earnings-related benefits.

4. Socio-political arguments. A society determined to care for people, and protect them from income shocks, is much more solidaristic. This serves as a protection against reactionaries coming into power¹⁴. As I’ve argued in the past, bringing the poor and middle class into coalition is crucial. The alternative leads to divides that are easy for reactionaries to exploit, especially along racial lines. Social Security and food stamps are both welfare-but even though Social Security is WAY bigger, food stamps are attacked more, because it’s seen as being for poor black people instead of everyone. This kind of universal approach works in other countries too to fight reaction¹⁵.

So, given these justifications, what is the best way to construct social insurance? Well, it should be universal, for arguments 1, 2, and 4. And, due to the paradox of redistribution and argument 3, it should be earnings-related. Due to 1 (general anti-poverty arguments) 2, and 3, it should have a minimum benefit so poor folks are adequately covered. Note that these last 2 don’t really apply to health insurance, since ultimately health insurance helps you pay for a service. The failures of private health insurance actually have many unique elements due to several things unique to it. So I won’t go on too long about it here, and will instead write an article later about it and private healthcare in general. In any case, social insurance should also come first and foremost from the national level, to pool risks better and redistribute across the country. Other governments are, of course, welcome to add to it.

Perhaps you’ll note that there is one important type of insurance I haven’t mentioned: life insurance. Let’s apply our arguments here, and see if any of them fit.
1. Losing someone of course causes hardship on families. Not just emotional, but financial too. “The average funeral costs between $7,000 and $12,000”. Dealing with this on top of planning the funeral, settling the estate, etc. is an absolute nightmare. So, I would argue the state should step in, and ensure basic services will be paid for. Families deserve a proper goodbye, and shouldn’t have financial concerns about getting it.

2. The state already deals with bodies that do not get proper funerals. It isn’t remotely pleasant, but it is cheap. You can read about some horrible experiences of folks who couldn’t pay for a family member’s funeral here. In any case, I don’t think there’s an economic argument to be made.

3. The average household spends $465 on life insurance a year. This is a mere fraction of overall spending, so redistribution isn’t much of a concern here. But it wouldn’t hurt, might as well, right? Not so fast! Life insurance isn’t just used to pay for funerals. It’s also used to settle debts, which can of course be large.¹⁰ Remember, social insurance should cover something everyone deals with. Everyone deals with death, but not with large debts, especially if we’re making college free and housing more affordable. It’s also true that life insurance payments are often large to ensure dependents are cared for. But with a proper welfare state, this won’t be necessary. All in all, best to keep it to a flat benefit.

4. Certainly, a society that takes respects the dead and the loss of their families is a more solidaristic one.

So, 1 and 4 apply, and we should have a universal cash benefit. There actually already is a lump sum benefit paid by Social Security, but it’s a measly $255, and only goes to people in the household of the deceased who are receiving Social Security benefits¹¹. Let’s drop that requirement, and give the money to the estate, heir, or whoever is applicable. And let’s increase the benefit to the yearly poverty level, at $12,760¹². This number is easy to keep track of, it’s important (I intend to tie other benefits and tax brackets to it), and it should cover the costs of a decent funeral, and then some to ease the loss. You may need to settle an estate, or sort through the deceased belongings.
There were 2,813,503 total deaths in the US in 2018, so this policy would cost about $36 billion a year. This is minuscule compared to the rest of the economy! So it’s not a big deal at all to pay for it.

Until next time, parhain terveisin toverit!



Jacob Keegan

Economics and welfare knower.