The most recent issue of Res Publica features a collection of articles on social insurance and the welfare state, a topic near to my own heart. It was recently featured by Bookforum under the heading The Greatness of Modern Welfare States. I thought I might say a few words on this paper (ungated here) by Xavier Landes and Pierre-Yves Néron (two of my former postdocs, I should mention). Much of the discussion is a response to this paper of mine (which is actually just a more academic presentation of an argument developed in my book, The Efficient Society), in which I basically present a philosophical defence of that standard “public economics” view that the major role of the modern welfare state is to correct various forms of market failure. Much of this activity gets misclassified, however, as “redistribution,” suggesting that it follows some sort of an egalitarian logic, when it fact it is just an insurance scheme being run in the public sector, and is therefore no more redistributive than any other sort of insurance. The superficially redistributive character (of, say, public health insurance, retirement pensions, worker’s compensation, etc.) leads people to overlook the efficiency rationale for state involvement in these sectors.
First off, let me say that I agree wholeheartedly with the discussion in section 3 of their paper, in which they point out the tensions between luck egalitarianism and the “social insurance” features of the welfare state. Here I think the crucial observation is just François Éwald’s point, that there is a contradiction between the “delictual logic” of classical liberalism, with its focus on personal responsibility, and the “assurential logic” of the welfare state. As his study of workplace accidents shows, the only way to establish the “worker’s compensation” insurance system was to persuade people to stop taking workplace accidents to trial, and simply to pay out compensation based on a fixed formula. In order to persuade people to accept this, you needed to get them to stop thinking of the issue in terms of personal responsibility (e.g. “whose fault was it?” “who caused the accident?”). Since luck egalitarianism places such huge emphasis on responsibility, it is not surprising that insurance systems will often work at cross-purposes with “equality” as luck egalitarians conceive of it. (I once heard John Roemer say that his version of luck egalitarianism could just as easily be described as a patterned system of distributive justice based on the principle “to each according to his level of responsibility.” Scales fell from my eyes.)
So that’s all fine. When the discussion turns to the “relational” conception of equality, however, I think that Landes and Néron keep their eyes too firmly fixed on the luck egalitarian position they just criticized, and so fail to see exactly where the burden of proof lies in defeating my own efficiency-based perspective. Crucial to my argument is the idea that a “normative model” must be able to explain how things came to be the way they are. This means that, with respect to insurance systems, it must be able to explain why some are in the public sector and others are private.
My central argument against the egalitarian understanding of the welfare state is that, if private markets for certain types of insurance were working properly, the state could satisfy the demands of equality just by giving poor people a transfer or a voucher – as many have proposed with health-savings accounts – so they could afford to buy insurance; it would not be necessary for the state to actually provide them with the insurance. So to the extent that the state does directly provide insurance, the reason for this must not lie in some sort of egalitarian concern, the more likely reason is that private markets are failing to provide that sort of insurance – either at all, or at the right price level – and so the state must go beyond just providing a transfer and must provide the actual product. The basic rationale then for state provision is an efficiency logic – it is responding to a market failure. Of course, state provision may have a number of desirable side-effects. For instance, by charging based on ability to pay rather than actuarial fairness, it is able to promote greater equality. But this does not provide the rationale for the program, since you could achieve the same equality objectives without state provision of the good in question.
So in order for “relational equality” to displace efficiency as a normative model of the welfare state, Landes and Néron must not only show that a particular program promotes equality in that sense, it must show that this is not just a desirable side-effect. In order to do this, they must show that this form of equality can only be promoted by having the program in question be delivered in the public sector. After all, if we assume that efficiency is not the rationale for having a particular insurance program in the public sector, then we are assuming that it could just as well be delivered in the private sector. Thus the “relational equality” model would need to tell some story, explaining why the private delivery of the insurance product would fail to promote that form of equality – and why this failure would be considered sufficiently troublesome to people that they would want to nationalize that segment of the insurance industry.
Landes and Néron present no argument to this effect. Indeed, the absence of an argument is what leads me to suspect that they don’t see – or don’t agree with me about – where the burden of proof lies. Here is what they say at the crucial point:
Public insurance schemes are more than strict risk-management devices. They create specific moral relations among individuals and can potentially send a powerful message of equality. Policyholders become part of a community of insureds: they face uncertainty together and are collectively liable for the losses experienced by any one member of the community. As such, insurance provides a concrete framework as well as a normative vocabulary that individuals use to address each other, and which determines the kind of arguments they may use and how they may use them. In other words, public insurance furnishes a social grammar that allows citizens to make equality claims. This grammar is visible in the various ways that claims for compensation within a public insurance system are framed and addressed to other policyholders. Entitlements to compensation no longer stem from charity or means-tested investigation. Such entitlements are the direct consequence of belonging to a specific community, the insurance pool—namely the state in the case of public insurance (149-150).
Everything that they say here about the “moral relations” established among individuals in an insurance pool is true of all insurance schemes, whether they be private or public (this is tacitly acknowledged in that last sentence, when they say “the state in the case of public insurance”). So when they say that “public insurance furnishes a social grammar that allows citizens to make equality claims” it seems to me a non sequitur. All they have claimed is that insurance furnishes this grammar – and according to them, the private sector is just as good at delivering insurance as the public sector! Thus the argument does not really provide any explanation for why public insurance is necessary in order to promote this specific form of equality. Whatever solidarity or moral relations are established by the public insurance scheme is best seen as an auspicious byproduct (something that I once mused about, with respect to Canada’s single-tier health insurance system), but not as the fundamental rationale.
Finally, mainly as an aside, I should mention that the idea that certain “moral relations” are established among members of a risk-pool strikes me as empirically quite dubious. In general, insurance tends to encourage a certain moral laxity, where people inflate their claims because “insurance is paying.” They do so because the harm is diffused over many policy-holders, and it affects only anonymous others. (On this, see interesting discussion in Johannes A. Landsheer, Harm ‘t Hart and Willem Kox, “Delinquent Values and Victim Damage,” British Journal of Criminology, 34:1 (1994): 44-53 at 51.) If anything, I would tend to think that insurance weakens the bonds of solidarity in society – something that we tolerate only because of the benefits that it provides.
It seems to me properly addressing the question of whether the provision of social insurance is best handled via the public sector or private sector can’t happen without discussing the nature of money and its relationship to the purposes of the state versus those of private commercial interests. Private sector commercial interests operate in a realm where individuals are fundamentally unequal because one’s ability to engage with commercial interests and have full agency is dependent on the relative size of one’s bank account. (One could say the system is based on the notion of one dollar one vote.) But how does one define legitimacy with respect to how an individual’s wealth was obtained? Is it ok that a rich mobster, lottery winner, tech entrepreneur and heiress should all have more and better access to insurance than the poverty stricken single mom, restaurant dish washer, graduate student or child of the bankrupt business person? Money is terrible at distributing justice. The state is not a commercial enterprise. It’s purpose isn’t to turn a profit. On the other hand, a commercial enterprise in a capitalist system is organized to turn a profit; that is the point. The last thing we need is a system of social insurance delivered by commercial interests. We are guaranteed to get a system where we pay the most and get the least. Worse yet is a private healthcare system in which “consumers” are encouraged to purchase healthcare services and procedures and health insurance while getting as little real coverage as possible. It’s perverse. (Though maybe not quite as bad as private prison operators lobbying for mandatory minimum sentences.)
Certainly, it is well past time to reexamine concepts like “efficiency,” which have a rather particular meaning when attached to the mysterious construct, “free markets.”
Broadly speaking, efficiency is the ratio of output per inputs, but the rub is whose outputs and inputs are to be considered. Since the free market model frames labor as a commodity and cost burden like others, neither human needs nor potential are fully served. This is a systematic deficit in free market thinking, unresolved by the additional calibration offered by “market failure” or “externality.”
At the very least, we have to unbundle notions like free markets and efficiency to see more clearly.
I look forward to reading “Efficient State” to see how Mr. Heath approaches these issues.
And I can see that Mr. Heath’s argument is precisely what a moderate party in the U.S. should adopt as its underlying philosophy but, alas, there is no such party in the U.S. Please come down South and found such a party!
Maybe I’m missing something or read this too quickly, but I thought a normative model was simply about how things should be – something you could propose to other reasonable people that they would, if acting disinterestedly and dispassionately, accept, whereas something that explains how something came to be is just descriptive. So, I think your efficiency rationale for public health/pharma/eyecare/car/workers comp/pension etc insurance makes a lot of sense as a normative model with a happy redistributionist byproduct, but I think the reason we have, public health insurance, at least, is not based in efficiency, but based on a desire to have fewer poor people die. After all rich people already had insurance and I suspect there was not a pressing push by them to pay for poor people’s insurance through the greater tax hit, which would likely outweigh their marginal benefit from cheaper insurance. So I don’t know if you can divorce the tax code from the insurance system. In any event, assuming the government will pick the most efficient choice is sort of begging the question, isn’t it? Sure, it could have just given them money, but it has done all sorts of things out of a redistributionist impulse that are not necessarily the most efficient – giving everyone cheaper electricity, not tolling roads, nationalizing airlines, etc. At a minimum, it’s an empirical question – we should review the parliamentary debates at the time to see what people had to say – and one that, I’d assume, is divorced from the normative question of what people ought to do. Anyway, I’ll go and actually read the source references and links now. This is an interesting topic, thanks.
Hi Adrian: The actual story I subscribe to is a sort of “cunning of reason” one. Social insurance programs (for instance) have typically been introduced for reasons other than efficiency — usually for some sort of broadly egalitarian reason. However, the fact that they are efficient is what allows them to survive the next electoral cycle, when a right-wing government that is not committed to equality comes along and cancels all the purely redistributive programs. Because social insurance is usually pareto-improving in practice, there winds up being no big constituency with much to gain from seeing it abolished. Thus is differs from a purely redistributive program, where there will always be a constituency which a significant economic interest in seeing it abolished.
Anyhow, the idea of a normative model it that it should be “reconstructive” in Habermas’s sense — it shouldn’t be just a free-floating justification, but should provide some insight into how things came to be the way they are.
This is a powerful argument about justifications for social insurance. However, it seems ahistorical, particularly considering the US context. The social movements pushing for social insurance usually argued for a broader social equality in which access to health care or other benefits played an important role. Given that universal social benefits enjoyed broad popular support, conservatives and business interests faced a dilemma. Should they oppose statist and socialistic measures, fearing further socialization, should they propose Bismarkian conservative reforms, or focus on developing private programs reinforced by government sanction? The balance of political forces has generated a hybrid model defended by moderate social democrats and moderate conservatives for different reasons. Obamacare is a relatively conservative approach to extending health care in form, but it involves a substantial redistribution of wealth. It is legislative sausage and not ideologically coherent.