Winner of the New Statesman SPERI Prize in Political Economy 2016

Sunday, 14 May 2017

Should we demand ‘fully costed’ programmes?

Chris Dillow says we should not, and indeed that journalists who constantly ask ‘where is the money coming from’ are pandering to the idea that the height of economic competence for any government is to balance the books. I think his argument makes some good points, but taken at face value it is untenable.

To see why it is untenable, imagine a political party that promised to increase public spending, cut taxes, cut back on borrowing and let the central bank control inflation. Should journalists simply let that pass, as if the government budget constraint no longer existed?

You could respond by saying that a government that promised the earth would obviously not be credible. The problem with that argument is that a majority of the British people recently voted for a plan that would damage trade with our largest trading partner and most of that majority still believed they would be no worse off as a result. It is part of a journalist’s job to remind the public that basic trade-offs and constraints exist.

But I think Chris is right about individual policy measures. It makes little sense to require that each item of addition spending is matched to a measure to raise additional revenue, because this is not how fiscal policy actually works in any country. Whether all taxes should be tied to particular items of spending (hypothecation) is an interesting issue well beyond the scope of this post. Given this is not how current fiscal systems work, journalists and politicians should not encourage a belief that it is.

But if Chris is right about individual policy measures, when do we get the discussion of the overall fiscal picture that I argue is necessary? The answer is a simple demarcation. If an individual spending minister or shadow minister is proposing a particular measure, don’t ask them how it will be paid for. Instead ask them whether that measure makes sense on its own merits, and why doing something else within that minister’s remit would not be more preferable. For example ask an education minister whether it wouldn’t be better to avoid coming cuts to school budgets rather than spending money on grammar schools or cutting tuition fees.

On the other hand, if the actual or potential Chancellor is being questioned, it makes sense to ask whether the programme as a whole would increase or decrease borrowing. Chris is right that all a Chancellor can do is plan for a particular level of borrowing, but that alone is insufficient grounds for not asking about their plans. Instead what it suggests is a good line of questioning for journalists: if the deficit unexpectedly increases/decreases what would you do? With any luck that sort of discussion could involve some macroeconomics that went beyond bookkeeping.

It is here that we can judge macroeconomic competency. In the current context, for example, any politician that fails to note that we are in a liquidity trap (interest rates are close to their floor and the central bank is increasing the extent of its unconventional monetary policy) and that therefore some temporary borrowing on current account would be a good thing is either not competent or is for some other reason still attached to austerity. Any politician that says we must target the overall deficit rather than the current deficit and thereby hold back public investment despite real interest rates being approximately zero is not competent.

A really intelligent way of helping the electorate judge these issues during elections is to enable the OBR to cost the programmes of the main political parties, as the Netherland’s fiscal council does. All it would need is a modest increase in resources for the OBR, which would be a small price to pay to improve the level of public debate. Ed Balls asked for that in 2015, but Osborne refused. It was typically short sighted, because at the same time he could have given them the remit to cost the implications of leaving the EU. That would have allowed the OBR to tell us that Brexit would cost the government around £15 billion a year (Table B1) before rather than after the vote. If the assessment of the economic costs of Brexit had come from the independent OBR rather than the Treasury, that alone might have been enough to change the result.


  1. “Where is the money coming from” is a perfectly reasonable question to put to a minister in charge of a spending department. That’s first because individual ministers are supposed to speak for government as a whole, as well as their own departments. Second, it is reasonable to assume the deficit is a given, or if you like that government as a whole has got the deficit right. Or perhaps the latter assumption should be made explicit when questioning ministers.

    On the assumption that government has got the deficit right, a minister who proposes spending £Xbn extra should be able to explain where an additional £Xbn of tax will come from. (Of course the latter “£Xbn of spending requires £Xbn of tax” idea assumes the stimulatory effect of the former will equal the deflationary effect of the latter, which is an over-simple assumption, but for the purposes of quizzing ministers it’s an assumption which will do.)

  2. I was expecting to see something about Simon Wren-Lewis's pet bird, but then I realised I misread the title of the blog as "Mainly Ma Crow" ...

  3. There exists no capability in current macroeconomic analysis that would help non-economists understand what they need to about the issues. Until there is something as useful and accurate as the present weather forecast which is broadcast on television then any expenditure on more bureaucracy to crank-out hopelessly dogmatic macro orthodoxy will make matters worse.
    Nice try Professor but your attempt to get even more government funding for risible macroeconomic jobs-for-the-boys has been seen through.

  4. Is it really possible to make a clear distinction between current and capital spending ? For example, if you build a new road, that would presumably be classed as capital spending. If, though, you carry out regular maintenance to extend the life of the road, this would probably be treated as current expenditure. This seems a bit arbitrary to me.
    Similarly, most education spending is counted as current expenditure, yet in other contexts it would be described as "investment in human capital".

  5. I am not intimately familiar with how "Washington works," but we have this exact problem with Congressional rules now. You can only add spending by cutting somewhere else. (Of course, adding to revenues by increasing the tax rate is off the table.) Similarly, there is a lot of talk about "not wanting to 'spend' money on infrastructure" because "it adds to the deficit." But done right (and yes, plenty of infrastructure projects can have limited long term economic benefit, especially if not done right), these projects have long term positive contributions to GDP (cf. Smith). Furthermore, in the US (I don't think this is the case so much in the UK), the federal government doesn't maintain a formal capital budget separate from the annual appropriations process, which distorts how projects are evaluated, scored, and funded. (Excepting the City of Washington, which has the same problem) State and local governments have a traditional process, using on a six-year cycle, with annual or bi-ennial review.

    ... or the failure to acknowledge that cutting back on IRS staff "costs" revenue, etc.

  6. Whilst agreeing that a country such as ours with a fiat money system which does not have the fiscal constraints on it as they do in the European Union, I really do get annoyed that we seem to apologise for people like Goerge Osborne who clearly understand this and yet pursue a mythological paradigm which deliberately transfers wealth and power to people that already have it, at the expense of everybody else.

    George Osborne and indeed Ed Balls are both guilty of the same crimes, they were both working to the same ends whilst each blaming the other for the consequences.

    Ed Balls was the worst shadow chancellor I have ever seen in the Labour Party, Brown sadly was the worst ever Chancellor as he was instrumental in deregulating the Financial sector in line with the Tories.

    They were all in it together and all blaming each other for the financial crash.

    If politics and economics do not work for people then it will never work at all. Career politicians serve no-one outside themselves and are a very dangerous breed of humanity.

  7. Can you really cost a programme? This implies some assumptions about fiscal multipliers which in turn implies some assumptions about humans behaving like econs, etc. All we can do is make rough estimates of expected growth but not how relatively minor policy changes might change those expectations.

    I think there are two questions worth asking:
    1. Is there a need for the investment? For example, do we think that health or education and therefore economic outcomes can be improved?
    2. And on the question of competence. What is your overcome macroeconomic strategy? For example, what debt to GDP ratio are you targeting over the next 5 years, and what is your strategy to get there?

  8. «That would have allowed the OBR to tell us that Brexit would cost the government around £15 billion a year»

    That seems a reasonable estimate but it is just guesswork in the end, because "Leave" in the end is a game changer.
    This brings about all the complications the Congressional Budget Office has to deal with in the USA, notably about "dynamic scoring", which it has rejected for a long time from both Democratic and Republican requests, adopting instead an "other things remaining the same" approach which is not enlightening.
    The risk, or rather the likelyhood, is that estimate on proposed legislation end up being precise but rather inaccurate.

    BTW, the estimate of a £15b/year cost for "Leave" might have swayed some voters, but it is quite small in the grand scheme of things: for those who think of the EU as the Fourth Reich it is a very cheap price to pay for independence, and for those who see EU membership as a tool by the elites to drive down wages, that the upper classes will lose £15b a year does not matter.

    And there is another problem with the OBR etc. doing full costing: most Economists are very reluctant to estimate distributional impacts, for some reason, and yet most government policies have primarily distributional impacts, more important than "full costing" or "benefit analysis". How to reconcile the importance of distributional impact with the reluctance to talk about it?

    1. "most Economists are very reluctant to estimate distributional impacts, for some reason, and yet most government policies have primarily distributional impacts, more important than "full costing" or "benefit analysis". How to reconcile the importance of distributional impact with the reluctance to talk about it?"

      Couldn't agree more. Back in the wild 70's we studied a topic called 'Welfare Economics'. It seemed to me at the time important - who benefits, how much, and at what cost to others? But in years of following political debates, which often centre around this very topic, all I have ever heard from economists is "It's up to society to determine the distribution of the benefits/costs of any political proposal. This is a real cop-out and unworthy of the intellectual ability of economists willing to forecast to 1/2 of 1% the quantitative impacts of policies.

  9. «A really intelligent way of helping the electorate judge these issues during elections is to enable the OBR to cost the programmes of the main political parties,»

    And here we go back to one of C Dillow's main arguments, that the press and leaders don't really influence much elections, and a few days ago he also reported:
    Sean Kemp says: “The vast majority of people don’t pay any attention to the general election whatsoever” (2’30” in). Yougov, for example, have found that only 15% of voters have heard the Tory slogan “strong and stable” even though its endless repetition has driven many of us potty.

    If this is a reliable picture, scoring by the OBR might matter those relatively few people who are inside the mediamacro bubble, but it is going to matter to the election result less than the press or leaders.

    Most voters might be justified in not following the arguments on general election because most of them have a few decades of experience and they know well that estimates and scores rarely survive long, they know well that as that wise man said, "events, dear boy, events".

    I think most voters are sadly resigned to "trusting" their political side leaders to react to events the way they would like to; that is I mostly agree with T Blair's perception that:
    people judge us on their instincts about what they believe our instincts to be

    However I think that there are benefits to scoring: they help in the long run voters to understand what the instincts of the parties are, and act as a matter of record for later re-evaluation. The extra cost of a few Economists is minuscule.

  10. I think that I disagree with your point about full costing. The problem is that as soon as you go down the 'fully costed' route you are buying into the ideas (a) that public expenditure is inherently bad and (b) that expenditure can be directly sourced from some specific income stream. Add to that the inherent uncertainty of almost all programme costs and you are really joining in a mug's game.

    So what's to stop politicians promising the sky (apart from the observed fact that it's actually a very bad way to win elections)? Rather than a pounds and pence cost estimate, the commentariat should develop a format to identify what the real, opportunity, costs of a policy would be. For example, a promise to reduce the average time spent waiting in A&E should be countered by asking how many more doctors, nurses or other resources this will entail. The response then should - or at least could - lead to a reasonable discussion about the number of doctors or nurses not currently employed or available from some other area. Just saying "this will cost £80 million which we will raise from increasing stamp duty" is not really useful.

    The key point, which you among others have been making for years, is that public expenditure is near cost-free when there are unemployed resources. I note that Krugman is now saying he believes those days are over in the US but I doubt that they are in the UK.

  11. Although theoretically i agree with a lot of this practically it totally untenable for the same reasons you write this blog. Potications do not engage with arguments or even agree to debate (see this election and TM). So not asking questions from spokesperson(minster) of party when the deem to turn up is clearly untenable in the present context. If they did engage with the rational arguments we would not be in this mess in the first place.

  12. But there is a way to combine the two. Require the spending minister to justify anything with present expenditures and future benefits according to an NPV which is itself related to the cost of borrowing over the period that the activity will operate) AND that the costs be the opportunity costs of program inputs -- lower in times of high unemployment.


Unfortunately because of spam with embedded links (which then flag up warnings about the whole site on some browsers), I have to personally moderate all comments. As a result, your comment may not appear for some time.