I foolishly promised a few people that I was going to write something about “Predistribution”, which would not normally have resulted in actually writing anything about it, except that Chris then wrote his piece and I felt I ought to enter into the lists on the somewhat more sceptical side. In as much as it isn’t just a bit of industrial policy combined with “all things bright and beautiful” (More education! More skills! But who will empty the bins in this hi-tech utopia and how much will they be paid and why?), predistribution appears to be, as Chris says, an attempt to make all sorts of regulations and interventions in the economy do the work of a redistributive tax and benefit system. I don’t like this idea, basically for reasons to do with the fact that even after it all, I’m still an economist at heart. But the fact that I don’t like it doesn’t mean, in and of itself, that it might not be the best idea going in Britain today – after all, all the other politically live proposals might be worse. Read on, for a discussion of all these issues …
Of Dogs And Frisbees
To set out the skeleton of my argument here, my main point is that I don’t really agree that a robust and sensibly designed redistribution through taxation does constitute scrabbling around playing Robin Hood, and I think it’s an acceptable answer to the question “but why are they poor?” to say “they’re not really all that poor any more once you take into account the generous state benefits and public services provided to them”. I think that this is an intuition that’s pretty hard to argue for or against – I can see the other side of the argument but hold to my own. So I’m mainly going to be arguing in instrumental terms here; I don’t think this means I’m particularly at cross purposes to Chris, because most of the points he is making in his NL
RP piece seem to me to also be basically instrumental and practical. But in so far as there’s a separate, ethical question of whether it’s better to have your money come to you from a social program or from some other arrangement I doubt I’m going to be able to address that.
And chief among these instrumental issues is something similar to the point made by Andy Haldane of the Bank of England in a very good recent speech on financial regulation – that a complicated system for reaching some presumed desirable end has a significant disadvantage over a simpler system, simply because it is more complicated. Not only do complicated systems have more tendency to go wrong, they tend to offer much more opportunities for people to game the system if they are willing and able to devote time and resources to doing so. Since, in general, the initial allocation of time and resources is part of the problem here, it strikes me that “predistributive” systems are always in danger of either generating a load of unintended consequences, or piling up massive administrative overhead costs in an effort to avoid doing so.
One doesn’t have to be an out-and-out Hayekian to believe that some schemes are much, much more demanding in terms of information than others. Anything which is attempting to achieve some form of egalitarian outcome by tilting the regulatory playing field of a large and industrialised economy is going to be much more demanding than a scheme aimed at achieving a broadly similar outcome by simply taking from Peter and giving to Paul. In general, if you think that the scales are weighted against the worst-off in society, the solution is not to have a man in every shop, putting his thumb on the other side of the scales. Since one of the key proposals in the Miliband speech appears to be an aim to make banking policy do the work of industrial policy to do the work of regional policy to do the work of housing policy, I think I’m on firm ground in saying that there’s a real issue of complexity that needs to be addressed .
The Strange Death Of Tax-and-Spend Liberalism
The thing is that tax-and-spend redistribution is something that’s much more amenable to cost-benefit analysis. We know a lot about taxation, its incidence and the rough size of the inefficiencies and deadweight losses of a tax and it is correspondingly much harder for those affected by a tax to bullshit about its consequences than those affected by a regulation. Added to which, the world contains a number of really good, more or less unambiguous success stories for tax-redistributive societies (the Nordic countries), and many fewer of successful regulatory states. Why isn’t tax-and-spend more popular?
I think Chris’s piece is right on this; gradually over time and without always necessarily intending to, governments have built up a political coalition against redistributive taxation. The increased use of the means test has established a cultural assumption that tax-and-spend is always taking from one loosely identified group of “us” and giving to an equally loosely identified out-group of “them”. A number of policies have had this effect, but two of them really stand out to me. First, the sale of the council housing stock. This left the stocks depleted and in short supply, meaning that houses ended up needing to be allocated on a very strict needs-based means test in order to avoid humanitarian disasters, ensuring that social housing was unable to play its historical role and setting in place a lot of the problems clearly identified in those areas (particularly in the South) which have seen increases in support for the BNP.
The other big policy disaster was the introduction of student tuition fees and the abolition of the maintenance grants. It’s quite hard to remember, but up until the 1990s, nearly every university graduate started adult life with a great big present from the tax system in the form of three years’ subsidised independent living and cheap beer. Talk about how to build a coalition! One of the greatest mistakes of the last thirty years in my opinion has been the tendency to assume that measures which take money off the middle classes through taxes and recycle it to them via benefits are pure waste and regressivity. In fact, things like child benefit and student grants were one of the main ways in which the middle classes were bought into the system.
Conversely, if you want to undermine a program, start by means-testing it - one can see this at work in the USA, where a key plank of the program to destroy Social Security rests on the perceived unfairness of Warren Buffett receiving it. The public retirement systems of the UK and America were both set up as social insurance schemes rather than tax-transfer schemes at least partly for this reason. One might call the process of taxing higher earners in order to provide generous universal benefits, which are then themselves treated as taxable income by a name like “re-redistribution”, if one was on the lookout for a catchy title.
Spending a dollar-and-ten by trying to make 80 cents do the work of a dollar
So how did means testing get so popular, given that it’s so inimical to the goals of egalitarian politics? Well yes, that question kind of answers itself, but I think there was also a significant element of false economising involved. Means-tested programmes always score much cheaper than universal programmes in budget calculations, partly because they focus the benefit on those who need it, but also (and this is often surprisingly economically important), by making the benefit difficult or unpleasant to claim, a means-testing regime will create lower take-up of the benefit among those who are entitled to it.
Obviously that’s a pseudo-saving rather than a real one - saving money by not giving people money that they are entitled to is clearly not an efficiency gain, and can often be actively counterproductive since the unmet need of the discouraged claimants is very likely to show up as a social problem somewhere else in the system and to be more expensive to deal with when it does. But it shows up as a saving in the budget analysts’ costings of any specific plan, and the saving can often be more than enough to offset the overhead cost (which in turn, because it is a public sector administrative cost, is often underestimated).
And so there’s a tendency (very visible in the New Labour years) for modernising social-democratic governments who are operating under an aggregate spending constraint (or who have chosen, for political reasons, to act as if they are) to proliferate means-tested benefits. They look like they give you more bang for the buck, and they allow you to kick things off with a bout of cheap populism (like pretending, for example, that university education is a subsidy from the working class to the middle class). They also play into a particular pathology of the business-school and economics approach to government - to try to do analysis of individual projects at a granular level, rather than looking at the progressivity or regressivity of the system as a whole. Famously, the USA has one of the most progressive tax systems in the world (mainly because it doesn’t have a federal value-added tax). The greater progressivity and egalitarianism of, say, Sweden, is all on the spending side. And a granular, cost-oriented approach to redistributive projects is, at heart, a “lump of government revenue fallacy”, presuming that the opportunity cost of some spending program is some other spending program (in the UK, usually the NHS).
This gets the budgeting process backward in my opinion; rather than deciding what needs to be done and then deciding how to finance it, the government regards its revenues as fixed and then decides what to do. I think that this is what creates what Chris calls the “scrabbling around” effect, but it doesn’t seem to me that predistribution really mitigates this to any very great extent; to the extent that the regulatory manoeuvres embody stealth taxes and therefore disguise the extent to which resources are being transferred, it might make it possible to politically sustain a somewhat larger “lump of government”, but it doesn’t really get round the problem that a lump-of-government model is a bad way to run a social democratic state.
A bad idea whose time has come?
Something like this has, of course, been mostly visible in the European countries which were hardest hit by the banking crisis. In Ireland and Spain, the cost of financial bailouts have been almost entirely met by spending cuts rather than tax increases. Once more, this compares poorly to Sweden, which financed its massive bank bailout program by raising the top rate of income tax to 58% and introducing a 20% VAT. But Sweden in the 90s was the kind of society in which this was possible - because there was a general culture of social solidarity and tax compliance. People were used to the overall budget and the overall funding of that budget as something that the whole Swedish population was involved in; if they had been used to thinking of taxes as something taken from part of the population and given to another part, presumably they would have had just the same reluctance to pay them.
But we’re not Sweden. This kind of worries me, because the other end of the continuum is Greece, a society where tax compliance and consent to government has almost totally broken down, and the economic consequences of that are pretty obviously very bad. And although I’m not sure that there’s a slippery-slope self-reinforcing mechanism, it seems clear to me that in the last ten to twenty years, the UK has moved perceptibly in the direction away from the Swedish model of budget solidarity. The fact that predistribution is considered to be the best that we can do says something quite worrying about the state of British political institutions. But maybe predistribution is the best that we can do. And if it is, there’s not much point in pretending that it isn’t. It seems to me that it’s an inefficient, limited way to make the UK economy a bit more redistributive, within tightly constrained limits, but if there is no support for a Nordic redistributive state then so be it. I do not really have much feel for electoral politics at the moment, but maybe this really is a bad idea whose time has come.
 If your answer to this question is something along the lines of “foreigners”, then well, come on, seriously. More sinisterly, I think the actual response would be (to put it somewhat bluntly) that the bins will be emptied and arses wiped by “those members of the working class who, through indolence, fecklessness or simple lack of noble qualities, unaccountably failed to respond to the treatment of the dozens of Sure Start programs and parenting classes we lavished upon them in the name of predistribution; they had their chance and blew it”. Predistribution, for all the Rawlsian elements that people are trying to see in it, is a child of the meritocracy-in-the-pejorative-sense and does not seem to me to have escaped these roots.
 And if there is, then there’s a problem here which is much wider in scope than just social-democratic redistributive states – workers’ ownership of the means of production is fine enough for workers, but modern society contains more and more people who can’t work because they are old, young or disabled. Chris pointed out to me on Twitter that there is quite a discussion of this issue in the Critique of the Gotha Program.
 And I would even argue that any progressive society ought to be regularly making positive progress in the standard of provision for those of its members who would simply like to make the free and autonomous choice to do something with their lives other than produce goods and services in the economy. Marx doesn’t have much time for this bunch, in the Critique of the Gotha Program or elsewhere.
 Of course, not all predistributive policies are subject to this objection on grounds of complexity, although I think it’s pretty indicative of the problems discussed in my piece, Chris’s and the links that basically all the practical proposals tend very think-tank-wonk. Two examples of simple and straightforward policies which would be aimed at altering the relationships in the economy in a predistributive manner would be the minimum wage, and strong trade union rights.