Public service reform: memo to Osborne

Dear George.
You claim that the Tories have inherited Blair's commitment to reforming public services. Good. But there's a problem here. Reform of the services requires the co-operation of those who work in them. And to achieve this, you're going to have to rediscover the classical art of  rhetoric - the skill of persuading people who disagree with you. This requires (at least) six things:
1. Stop talking about "waste." Of course, there is huge waste in public spending. But saying so merely preaches to the choir. It doesn't win much support from public service workers. Nor, for that matter, does your talk of "lurches of the left" and personal attacks upon deputy leadership candidates. This is the tribal politics that alienates neutrals and provokes opponents into circling their wagons. But it's their support you need.
2. Talk more about equality. The great under-rated failing of the welfare state is that it's failed the poor. The focus of welfare reform should be on the question: how do we change this?
Of course, this takes substance, as well as rhetoric. Perhaps you should follow the examples of Texas, and equalize exam results across schools, thus ensuring no-one suffers from going to a bad school. Or you should ensure that poor families get the first pick of schools, perhaps by giving them more vouchers.
3. Empowerment. It's a basic principle of the economics of property rights that the party who owns the most important resource should control the organization. This means that organizations where human capital is important should be controlled by workers. This is why law and accountancy firms or vets' practices are partnerships. The same should be true for hospitals and schools.
Welfare reform should therefore focus upon empowering teachers and doctors. "Privatization" is a red herring; it replaces one bunch of superfluous managers with another.
In this sense, good economics combines with good politics - it'll be hard for unions to oppose measures that empower their members, though they'll try.
4. Let's hear more about diversity and less about choice. In your recent speech, you discussed both. But "choice" doesn't appeal much to public sector workers. Some think  patients and parents aren't informed enough to make choices. Others are persuaded (maybe too much) by Barry Schwartz.  Talking more of diversity of provision (or of control structure) would hold more appeal; let choice follow.
5. Make reform positive-sum. It's a fact about any policy that losers make more noise than winners. Nye Bevan had the solution to this: he stuffed doctors mouths with gold.  This, of course, means there's a trade-off between controlling public spending and reforming services.
6. Look for precedents.  Everyone has a status quo bias. So examples of how diversity of provision works in social democratic countries - Sweden's school vouchers, France's health care - are important.
Little of this, I fear, will come naturally to you. Which raises the question. Are you serious about public service reform, or are you just using this as a stick with which to beat Gordon Brown or as a disguise for the pursuit of traditional Tory vested interests?

Comments

[Shuggy]

"Talk more about equality."

Do you think he's likely to heed this advice? He's a Tory - he doesn't believe in equality.

[Thom]

George Osbourne came to my office last week. Fortunately I was out, but he ended up using my desk for the afternoon.

I've burned the chair, and salted the earth around it, but I still feel a bit dirty.

[Mark Wadsworth]

Good stuff. Have you emailed it to him?

[Heraklites]

"Reform of the services requires the co-operation of those who work in them."

If the problem is that services reflect the interests of suppliers rather than users, how is such co-operation to be got?

[Matthew]

Chris, I'd be interested in your response to the objection that was raised to this plan

"Perhaps you should follow the examples of Texas, and equalize exam results across schools"

that employers and universities will inevitable ask 'which school', and therefore discount good exam results from the worse schools (which in a sense would be correct).

[dearieme]

"Reform of the services requires the co-operation of those who work in them." Only if you stop short of just abolishing those services. If the Welfare State has failed The Poor, what's it for? Churning money wastefully among the mildly prosperous is a fool's game.

[Paulie]

That is a spectacularly good post Chris, even by your standards. Your conclusion, by the way, is spot on. The Tories aren't market socialists.

I disagree with one line:

"...it'll be hard for unions to oppose measures that empower their members, though they'll try."

It won't be hard for Unions to oppose such measures at all. Unions do not reflect the perspective of their members. If Unions were highly responsive democratic organisations, it would be hard for them to oppose such measures. But they aren't. They spring from the British tradition of (mostly) Christian socialists that prefer nationalisation because it is the only way to make the boss behave nicely. The (British christian socialist) tradition of mutualism has never been a dominant force in Trade Unionism here. In my experience, Unions define themselves partly in opposition to co-operatives. Staff in co-ops don't bother with Unions most of the time - why should they?

[james C]

Dear Chris,

I don't do anything unless Dave tells me.The donors tell Dave what to do.He tried a policy of his own on grammar schools, but I think he has learned his lesson. So it's back to saying nothing until Brown loses the election.

If you have £50k to donate, I will give your ideas some consideration.

George

[dsquared]

[that employers and universities will inevitable ask 'which school', and therefore discount good exam results from the worse schools (which in a sense would be correct]

just to add - statistically correct but individually potentially quite horribly unfair; the brightest kid in Knowsley Comp will have his achievements utterly devalued by the average for that school.

[Kevin Carson]

Re your third point, it seems to me that labor is almost always the factor with the highest monitoring costs and agency problems, even in the most capital-intensive firms. Residual claimancy by workers would probably be the most rational form of organization even in caital-intensive production. What actually happens is we start with the fact of structural biast toward capitalist ownership, and capitalist owners try to make an inherently inefficient form of organization stumble on with the agency problems minimized. To paraphrase Drucker, the bureaucratic corporation is the most efficient way of doing an inherently inefficient thing, given the fact that most people working in the organization have no rational interest in its goals. In fact, capital-intensive modes of production are themselves often such a means of adapting to the irrationality of the capitalist-owned corporation, a way of reducing the agency problems of capitalist ownership by substituting capital for labor. So it might well be that, for a manufacturing firm owned and managed by workers, the most efficient mix would be less capital-intensive.

[Sam]

the brightest kid in Knowsley Comp will have his achievements utterly devalued by the average for that school.

It is probably also the case that the results of children from small, unknown schools will be devalued compared to those from known good schools.

In general, I don't think it's much of a problem - exam results are a very blunt tool for selecting good pupils, employees or students. On the other hand, playing these games with people's marks only has the effect that Chris wants if you just trust the blunt tool, rather than interviewing candidates.

At the best, these games will get a few bright but poorly-taught kids past an initial coarse marks-based screen and give them a chance to show their worth, and deny the same opportunity to a few well-coached dullards.

I can't get too excited about that.

[Meh]

Dear Chris,

Some random points to consider:

- The French Health System is tottering under massive debt. To the point where they are gradually shifting towards a Danish (and even in parts a UK) model to control costs.

- Markets don't provide diversity. So United Healthcare won't be supporting your proposals.

[Peter]

Nothing like enough positive about how to improve things, yet. The Nordic model of politics and the public sector certainly works better, but so do things in Germany. Even the French provinces do better, although I’m told that it can be tediously bureaucratic. Certainly much from labour relations and management methods in the private and neo-public sector (the larger organisations in the not-for-profit sector) should be impressed on much more of the English public sector: choosing the right people for the job, training, formal quality management, regular performance appraisals, audit of new systems before they go public – but that has all to be part of empowering, enabling, and ensuring that the policies are delivered. That of course will temper the tendency of politicians to demand that specific changes be put in place by specific dates: where new systems, new groups of staff are needed, the necessary time must be allowed for development. But there is another factor in the public sector: personnel who are following their own personal agendas, so that teams of service delivery people are diverted from the expected method of delivery – but don’t let local innovation be mistaken for disloyalty.

[Luis Enrique]

If the equalizing exam grades idea is perceived to achieve what it aims to - identify smart kids from crappy schools, and down weight mediocre kids from good schools - them employers/universities will not completely devalue exams from bad schools will they? If you want to assume rationality on the behalf of profit seeking firms (or ability seeking universities) where recruiting smarter workers/students is the goal, then they'd be gains to be made from recruiting smart kids from crappy schools, if other employers/universities make the mistake of devaluing those grades.

If you think people people will just identify schools and adjust their valuation of grades accordingly, thereby canceling out the effect of grade equalization, why don't they apply weights already? Do universities and employers do currently downgrade the grades of kids from good schools and up weight grades from bad? If not, why do you think they're going to start to apply weights after the grade equalization is introduced?

If your objection to this idea is that rational employers and universities will devalue grades accordingly, then you have no objection other than to say that the reform will be neutral - you are arguing that people will not pay attention to the nominal and look at the real - well if they're capable of doing that, then they're doing it now. If they're not doing it now, they they do pay attention to the nominal.

If you don't like the idea of rationality, well then the scheme works because people care about the nominal. Either in procedural form (universities are forced - do whatever practicable degree - to accept students according to grade and are not allowed to apply subjective weights) or that employers/universities will pay attention to the nominal grade and not adjust fully, in which case the idea achieves its stated aim.

So either it does nothing, or it does good. I don't see how it does harm

(the case where very smart kids from crap schools are already getting 15 A* so grade equalization won't help them is not a worry, those kids are going to do OK).

[dsquared]

[If the equalizing exam grades idea is perceived to achieve what it aims to - identify smart kids from crappy schools, and down weight mediocre kids from good schools - them employers/universities will not completely devalue exams from bad schools will they? ]

But it won't do this and it doesn't aim to; Chris's whole intention here is to break (or at least weaken) the connection between individual ability and the grade given. So it doesn't actually identify "smart" kids at all. If we say that I get 60% in the exam and Chris gets 80%, but I go to a crappy school and he goes to a good one, then it could easily be the case that I get an A and he gets a B. If something like this happens every year then it isn't a mistake for employers to devalue my school's grades; if they don't, they'll end up with loads of people like me and none like Chris.

[Do universities and employers do currently downgrade the grades of kids from good schools and up weight grades from bad? If not, why do you think they're going to start to apply weights after the grade equalization is introduced?]

Universities don't because A-level grades are currently given against an objective scale. Employers certainly do downgrade the grades of graduates of bad universities - try getting an interview with a 1st from South Bank University versus with a 2:1 from Cambridge. Because employers do this, I would guess that universities would do the same if A-level grades were made more like degrees.

And this is how it can cause harm, because the downgrading is a blunt instrument applied across the board; I happen to know at least one guy who is a really, really good economist, but who is basically screwed because his degree is from South Bank, and all his potential employers regard SBU degrees as worthless. Something similar would happen under Chris's scheme almost certainly; you don't help bright kids by taking away their only method of signalling their brightness.

[Luis Enrique]

Yes, grade equalization will dampen the signal from very smart kids at bad schools, but enough to make a meaningful dent in how many of them get to a good university? Or at least enough of a dent to offset the potential benefits to less smart kids?

Only the very smartest would experience no uplift from grade equalization and only the down weighting effect, but when you're that smart, you're likely to get in to a good uni anyway aren't you? Less smart kids from bad schools will see their grades raised and then reweighted. The point of this scheme is surely to give the rest of the ability a distribution a leg up, not the super smart.

I don't know whether you are right to say that the point of this scheme is to weaken the link between ability and grades. Isn't the idea based on the premise that for any given ability your grades will be higher at a good school than at a bad one, hence kids at bad schools are at a disadvantage and the link between grades and ability is biased against them.

So the by adjusting for the effect of being a bad school on grades, the scheme is trying to tighten the link between grade an ability (OK given that 'true' innate abilities at bad schools are likely to be lower, then it goes further than that and introduces nominal bias in the other direction).

If you don't believe that it's harder for kids to get good grades at bad schools, then there's no bias to correct and the scheme is as you describe it. Perhaps you do think that; if I read you correctly you argue that universities do not already apply weights according to school because "A-level grades are currently given against an objective scale". So does that mean you think somebody with three B grades from Eaton is objectively equally as able as a kid with three B grades from Knowsley Comp? I'd be surprised to learn that.

[dsquared]

[So does that mean you think somebody with three B grades from Eaton is objectively equally as able as a kid with three B grades from Knowsley Comp?]

Yes. In fact, Oxford University specifically has a policy of mentally "upgrading" equivalent grades earned at worse schools. Which would obviously have to be scrapped under Chris's scheme.

A more sensible scheme would be what they actually do in Texas (I am pretty sure Chris has got the wrong end of the stick wrt the Texas scheme if he means what he wrote here) and say that the top x% of every school have to be given a place at a good university no matter what grades they get. This has most of the advantages of ironing out the differences between schools, without the disadvantage of making the grades system meaningless.

[Luis Enrique]

I'm confused! "Yes" you say, these two B graders would be of objectively equal ability - but then you say Oxford mentally "upgrades" grades from worse schools - which suggests to me they think grades are not objective measures of ability. So although you answered "yes", Oxford would answer "no".

Such mental adjustments would not be scrapped under Chris's scheme - they'd just change sign - if universities are already (fully) adjusting to separate the real from the nominal, then Chris' scheme would have no effect (this would also imply that kids from bad schools are not at a disadvantage getting into uni as things stand). If universities are not fully adjusting then the bias is still there and Chris' scheme may have an effect.

But sure, the way you describe the scheme makes sense. Chris' version would have interesting dynamic effects (would parents start sending kids to bad schools to increase their chances of getting better grades?) and also it would tricky to measure school performance if all schools posted the same distribution of grades, so would still need other measures, I think.

[dsquared]

No, they've objectively achieved the same test scores. The kid who did this with less help might have greater ability or he might not. Under Chris's scheme, though, it would be more likely that of two kids with grade B, the one who went to the worse school had got more answers wrong.

[Luis Enrique]

Yes, now you've got it! That is indeed the current system, and Chris' scheme would indeed change things to effectively raise the grades of kids with more marks wrong, from worse schools.

But (and isn't this what we've been discussing all along?) the current system constitutes a biased measure of ability, because some kids with the same number of questions wrong went to a good school and some to a bad school, hence Oxford's mental adjustments and hence Chris' proposed scheme (which may or may not work as intended, as discussed).

Although perhaps you don't buy the idea that there is a systematic bias from objective grading against kids from bad schools, because you wrote "the kid with less help might have greater ability or he might not" rather than "probably does have greater ability". But as you point out, Oxford for one does see a systematic bias.

[dsquared]

But how on earth can destroying the information in the grades be the best way to deal with that problem? At present, you can at least adjust the grades. Under Chris's proposal this is much more difficult, and the heuristic will be to assume that an "A" at Knowsley Comp is worth much less than a "B" at Eton. This is, as I keep saying, what actually happens with university grades.

[Matthew]

You'd have to have two grades - one the actual and the other the adjusted, and make it illegal for universities to discriminate on the unadjusted grades. Then at least you would keep the information from the actual grades.

I'm not necessarily against the idea, but it'd surely be politically impossible too, especially for the Tory party? If (say) 10-15% of pupils a good private school have to get ungradeds, Es, and Ds, regardless of their objective standard, it wouldn't be very popular, with many Tory MPs and voters?

And would it affect the incentives for schools to improve?

[Sam]

dsquared:

...at a Texas state university, which the state of Texas can do because it owns the universities.

Being the smartest kid in a bad school isn't going to help you get into Harvard.

[Andrew Duffin]

Just a couple of simple thoughts.

It's true what you say about law firms and accountancy practices, but it doesn't really transfer to schools or hospitals.

In the case of hospitals, the human capital is highly significant, but it's ineffective unless backed up by truly massive financial capital and enormous support structures. This isn't the case with (say) lawyers, who really just need a desk and a telephone, so your analogy doesn't really hold.

In the case of schools, the human factor may well be dominant, but in the special case of state schools, they have been so much captured by producer interests (the bureacracy, the teachers's unions) whose aims are not education but rather credentialism and political empire building, that again, I think it breaks down.

[Gadgie]

The statement that the welfare state has failed the poor is simplistic. It has been a success for many now affluent people who would have been poor without it but are no longer dependent on it. At best the statement should read that it has failed the residual poor. That is, of course, a truism because if it hadn't failed them they would not still be poor. Pedantry I know, but an absolutist approach that the welfare state is a failure is only made possible by ignoring where it has been successful. Of course, some of the poor may also be transient poor, about to use the welfare state to escape poverty. I work in adult education and so have seen hundreds of people use benefits and subsidised education to radically alter their lives. Many would still be locked out if this had never existed.