Lowering the NAIRU

Among the relatively few points the opposition has scored in this Parliament has involved the unwillingness (or, in the Opposition’s telling, inability) of Treasurer Wayne Swan to respond substantively to a question from Malcolm Turnbull about the level of the NAIRU (non-accelerating inflation rate of unemployment), AKA “the concept formerly known as the natural rate”.

At this point I was going to refer readers to Wikipedia but, as with quite a few economics articles, it’s not entirely satisfactory. However, rather than complain, I’ve edited it to include a slightly better explanation.

Coming back to Australia, the fact that inflation is rising suggests that, if the NAIRU exists, we are now below it. It doesn’t seem as if there is much scope for fiscal and monetary policy to be tightened further. Given the risk of a breakdown in global credit markets, raising interest rates any further seems very dangerous. And the tax cut promises (which should be kept – the credibility of political processes is more important than the risk of inflation) mean that the scope to tighten fiscal policy is limited.

What remains is the possibility of reducing the NAIRU by improving the performance of labour markets. Education and training will help in the long term, but not so much in the short run. What is needed is to take advantage of the tight labour market to reduce long-term unemployment and to bring discouraged workers back into the labour market. At this phase of the cycle, the best policy instrument to achieve this goal is a targeted wage subsidy. Employers who take on workers moving off unemployment and disability benefits, or re-entering the labour force after a long absence should receive a subsidy for a period of say, three to six months. I’ll try to post a bit more on this, and why it’s superior to suggested alternatives like cutting minimum wages, before too long.

96 thoughts on “Lowering the NAIRU

  1. Confucious says: When a mouse grows at 10% per annum indefinitely, your children will have an elephant in the house. When that elephant grows at 10% per annum indefinitely, your grand-children will have no house.

  2. Wage subsidies have been used very effectively in the past and the Job Network can still use this method. The reason that it went out of fashion is that it eats into the bottom line.

    The Job Network has payment parameters which operate against meaningful training and wage subsidies. It is one of the reasons that we have skills shortages as employers have been reluctant to train and employees are often too insecure, poor or uneducated to pay for their own.

  3. Granted that some (many?) long term unemployed end up on DSP. Nonetheless, ABS statistics show that long-term unemployment and very long term unemployment has been consistently falling for the past 14 years. There were no wage subsidies for most of that period.

    If it wasn’t needed then, surely it isnt needed now.

  4. So much has been written on this thread that it has taken me this long to compile a compendium answer addressing matters arising. I only hope that things haven’t gone stale in the interim. This mostly follows particular comments:-

    Terje commenting at 6, the Swales method achieves the effect of relaxing minimum wages by still delivering them while getting rid of their marginal cost to employers. It’s not a true subsidy since there is no funds outflow, unlike other similar suggestions.

    Mugwump commenting at 7, Ian Gould’s point commenting at 8 is partly right, since negative income tax does have a large funds outflow until wages paid adjust downwards enough for enough people to price themselves into work (after which tax revenues go up to cover the outflow). Basic Income/Citizens’ Dividend schemes are even worse that way (this is the problem with the suggestions of Terje commenting at 34). However, they are all long run equivalent to the Swales method, particularly if that is implemented with anonymous transferrable vouchers since those will eventually get monetised. Ian Gould is missing it when he says “how about how the whole point of the proposal is to increase the labor force by encouraging employers to give the long-term unemployed a go?”, because a broad based system reaches those too; targetting is what makes schemes like this fail, because it mostly just moves unemployment from one group to another (I won’t go into the games theory behind this).

    Uncle Milton commenting at 12, you are almost right about the lack of skills stopping employers hiring the long term unemployed anyway, subsidy or no subsidy. However, broad based continuing schemes make it cost effective to take them on and train them, plus they plug the leaks that keep topping up the pool of unemployed, which means that new opportunities do go to the existing pool – there is a cumulative improvement. This may be what Salient Green commenting at 15 has in mind.

    Hc commenting at 17, there is a problem with “…the correct logic of [JQ’s] position. Abolish all minimum wages, fair pay commissions, the lot and allow markets to determine wages for all. Then boost those wages using transfers for all those society believes are underpaid.” It’s from the sequence and the time lags. Between doing that and improvements coming in, you don’t just have large funds outflows with no matching tax revenue gains, you get a lot of human suffering. It’s like releasing a tourniquet before you solve the underlying problem the tourniquet bought time for.

    Ernestine Gross commenting at 23, with all due respect, you are mistaken in thinking that “It is exactly the targeting of specific actual and possibly short-term problems that is called for”. That approach stops certain immediate symptoms but rearranges and perpetuates underlying stuff. That is the tourniquet difficulty; but we have tourniquets in place now. Kim Swales’s modelling (op cit) has shown how the broad based continuing approach applies to a wide range of economic conditions, and typically boosts GDP by about half the proportional increase in employment.

    Backroom Girl commenting at 29, you are right in thinking that “…employers aren’t really interested in them unless they are broadly enough targeted to enable them to employ someone they would consider employing anyway”, but wrong in thinking that “the problem with wage subsidies that they are only really cost-effective if they can be fairly closely targetted to people with a low probability of otherwise getting a job…” [emphasis added]. They are only cheap enough that way, but they are actually less cost-effective because they are not very effective. As I repeat, the Swales tax break approach is not a true subsidy and costs nothing (what happens is, tax revenue falls as the unemployed numbers fall – but Social Security outgoings fall in step, budget neutrality not revenue neutrality). This also covers the objections of Bring Back CL’s blog commenting at 32.

    Terje commenting at 34, you are right about how to address inflation. However, the employment connection still matters. Since the subsidy approach is a more or less independent policy option, using it frees up more of the economic levers to get at inflation.

    Salient Green commenting at 43, you have put your finger on something the Swales approach does not address – how to stop economic activity from moving completely out from under the tax base anyway? Well, the long term answer involves changing the corporation tax approach to one getting revenue from shares, so the revenue keeps coming in anyway, and making certain other tax reforms I have neither time nor space to go into. However, there are serious political problems with that, stemming from sovereign risk (i.e., you can’t trust politicians not to change the rules on you). We would have to make constitutional changes to keep them honest.

    Terje commenting at 45 (and commenting at 48), you are wrong about “their loss is our gain” in “If foreign nations wish to subsidise their producers then it is certainly stupid and it isn’t free trade but their loss is our gain”, in a couple of important ways. You are using some aggregate “we”, and there are actually some real forms of damage that can be done that way, even in aggregate, notably by damaging going concerns, effectively throwing away the investment represented by the fact that they are going concerns and not merely a collection of passive assets. That’s why it’s a bad idea to close down less profitable mines when slightly better opportunities are found, say.

    Ian Gould commenting at 52, it’s true but misleading to say “it’s as near to a universally accepted proven fact as you’re ever likely to get in the social sciences that free trade increases incomes for all participants and that the countries with the lower wages pre-trade liberalisation gain the most from free trade”; “all participants” actually means all countries taken as a whole, not all economic players. Furthermore, many things are deliberately misclassified as “free trade” when they are no such thing, and actually harm those on the receiving end – notably “capital transfers” into countries that are wholly or mainly fiat currency flows actually represent the mobilisation of local resources, not any inwards flow of physical capital, yet they must be serviced just the same. But we are getting off topic.

    Derrida derider commenting at 62, you shouldn’t dismiss Salient Green’s account like that; you are distorting the data to fit your theory, in what sounds very like accusing him of false consciousness. Rebut him or bring in other data, don’t “correct” him.

    Fred Argy commenting at 70 brings in points that are important, but different from the topic. The point of connection is that addresing unemployment separately allows a freer hand in addressing these things. (Fred Argy had a letter on these matters in the Australian Financial Review of 22.2.08; I submitted a follow up, which I am pasting in at the end in case it doesn’t get printed.)

    Jill Rush commenting at 77, I’ve covered the costs of subsidy approaches and the training issues further up.

    E.D. commenting at 78, “If it wasn’t needed then, surely it isnt needed now” is the “syphilis goes away” approach. It was needed, we just didn’t have it, and this isn’t an either/or thing anyway – we can do these subsidy approaches as well as anything else that we did that worked, or as well as any outside good luck that might bail us out.

    Here is the body of a letter on this area I have just sent to the Australian Financial Review:-

    Fred Argy states that while many economists have worked on ways of attacking unemployment while keeping inflation from rising, few have found fast working ways.

    One of those who has is Professor Kim Swales of the University of Strathclyde. Where the “five economists’ plan” works through changes to the income tax structure, his approach has a direct impact on employers, offsetting their GST by a fixed amount per full time employee on the payroll (pro rata for part-timers). As such it is far faster acting on unemployment, and it is budget neutral with no funds outflows even during the much shorter lags since Social Security outgoings fall identically in step with unemployment. As the Swales plan is independent of other policy measures, it is practical to use those to keep inflation down even while unemployment is improving – but a full coverage of available options would require more space than a letter allows.

  5. Drat, serious finger trouble. I meant to put:-

    Here is some material from Paul Craig Roberts that covers outsourcing issues and some things that can be done about their employment consequences.

  6. PML,

    Trying to get my head around the Kim Swales proposal.


    From what I can tell it entails:-

    1. An increase in the rate of GST.
    2. A company based rebate on GST where the rebate is a flat dollar figure times the the number of employees.
    3. An initial state where 2 cancels out 1 in terms of the government budget position.
    4. Over time 1 remains the same but 2 grows because employment will grow.
    5. The growth in employment stems from a tax bias against capital in favour of labour.

    In essence it seems to entail a basic income paid to the employee via the employer. Have I got this much correct?

  7. Terje, you are almost right – pretty much right for most practical purposes. However, some of the technical details do make a difference.

    I hope this clarifies things:-

    – You are right about 1, the way Swales put it forward. However, it’s equally practical to change other parts of the tax system to keep tax revenue unchanged between day -1 and day +1 of implementation. You don’t have to raise GST rates, it’s just more convenient (and you don’t actually have to use a GST as the carrying tax, it’s just the broad based tax with point of impact on employers that we’ve already got in place – at state level, payroll tax could be used).

    – 2 is almost right, but small businesses get into the act too, not just formally organised companies (but this is probably just a quibble). Another quibble is that it’s not really a rebate but an offset, since it only gets used in working out tax due and doesn’t actually get paid out by the ATO (someone else told me that once). Also, the precise size of the rebate/offset needs to be spelled out; initially it should match Social Security costs and on costs (and it should be fixed in dollar terms to provide an automatic stabiliser as it adjusts in real terms over time – but that’s another story).

    – 3 is stronger than “An initial state where 2 cancels out 1 in terms of the government budget position”, it actually cancels out in terms of the tax revenue position. It gets clearer if you put your points 2 and 3 the other way round.

    – 4 is accurate but incomplete. It really needs to be read along with “…budgets aren’t affected because Social Security outgoings drop in lock step”. Over time, you get budget neutrality.

    – 5 is accurate in relative terms, compared to what happens now, but it’s actually describing underlying stuff the wrong way round, like early scientists guessing whether electrons were going to be positive or negative and picking the wrong one. What is actually happening with the Swales method is, undoing a tax bias that encourages lower employment since each retrenchment spreads the cost of unemployment over the whole tax base and not just the downsizer. So it’s removing a tax bias that favours capital over labour, not putting in a new bias. The reason this matters is, putting in a bias lowers GDP and someone somewhere has to end up worse off; removing a bias actually gains you GDP – and Kim Swales’s modelling supports that this would happen.

    I hope my publications page brings out some of these things clearly enough – feedback on that would be welcome.

  8. Okay there are several questions/concerns but the key one at the moment is that I’m not clear on why you think downsizing a company currently imposes a biased abount of cost on the downsizing company. Or are you saying that it inflicts an unreasonable or biased amount of cost onto the worker that is made redundant?

  9. When a firm downsizes, the costs and benefits it has to take into account are what each worker can do for it and what it costs to keep him or her on the payroll. Once he or she is downsized, Social Security outgoings go up – but that cost gets spread over the whole tax base, so it doesn’t get factored into the firm’s decision (or the other way round, when it hires). Countries that don’t provide enough Social Security face “Vagrancy Costs” instead, which show up in the cost of policing etc. Whichever, there are real costs that the firms spread, and that’s the bias we have now. The Swales approach is a “Pigovian” fix to this market imperfection.

    Longer term, it would be better to transition to a Coasian fix with no government involvement, but the Swales approach is the one that fits present circumstances better because it is fast acting and a good match to what we have in place already, like GST.

  10. but that cost gets spread over the whole tax base, so it doesn’t get factored into the firm’s decision

    I’m not sure why it should. I tend to see it as a passive decision rather than an active one (ie I won’t decide to continue employing you next week) and as such I don’t see it as the firms cost to carry.

    However given the existance of social security it is clear that sacking somebody who then goes onto social security will impose a cost that is spread across all producers.

  11. To #79

    We had wage subsidies in place in the 80s but not in the 90s. LTU fell during both periods, so wage subsidies are not required.

  12. E.D., the conditions in the labour market were very different during the 1980s from those at present. There are 2 sides to the labour market, demand and supply. In contrast to the 1980s, at present it is a ‘tight supply’ which is alleged to be the problem. The idea of a wage subsidy in JQ’s post, as I understand it, is to lower the NAIRU (assuming it exists), by means of public payment for adjustment costs (wage subsidies)or to top up ‘internationally competitive’ wages which are too low to entice some people into the labour market. The suggestion that it is a short-term measure allows for the possibility that the financial asset value deflation, which is in progress since August last year, will reduce excess demand in the labour market and the life of the commodity boom is finite.

  13. Terje, your last paragraph at 86 should mean that you now see that the Swales method in fact undoes a market imperfection. But your second last paragraph can mean either of two things, depending on whether you are talking “should” in an engineering sense or an ethical one.

    In an engineering sense, the Social Security costs “should” reach the firm precisely because otherwise it is passive – nobody ends up connected to it, the costs get spread, and you get a market imperfection. You “should” always make costs connect to those who affect what happens.

    In an ethical sense, there isn’t any ethical duty for the firm, yes – but neither is there for the tax base, so there is no ethical guide if those are the only two possible places to put the cost; the unemployed can’t carry it, as things now are. Ideally, people would have enough independent resources of their own so they could price themselves into work and still survive without outside support – but it’s not like that so the Vagrancy Costs issue comes up, which isn’t an ethical problem at all. Something has to be done about that or suffer the consequences, and here and now we have Social Security for that – which brings us back to the practical problem, which has nothing to do with ethics. If you want an ethical side to this, look into the thinking behind Distributism, which aims at people having enough independent resources of their own.

  14. Terje, as I mentioned before, in the long run a Basic Income system, Negative Income tax and the Swales approach at the same levels are all equivalent (actually, a Basic Income has to be set below comfort levels to work properly, or people won’t need and go for top up paid work and the economy will falter from putting in a bias the other way than we have now – most schemes aim at setting it too high like that). The catch is in how quickly they take effect and how much they cost until then and after that; I have listed them in increasing order of speed and cheapness (the Swales approach costs literally nothing over and above what is already being spent on Social Security anyway). On the ethical side, a Basic Income funded in the conventional way via tax revenue has even more ethical implications, from how it is paid for, but it is a step in the direction of everyone having their own resources; you would get that from a Basic Income funded from a special fund, drawing on revenue yielding assets like shares, and the final step would be to parcel that out as individuals’ private property with some arrangement to stop that being dissipated, particularly over generations. But that brings up other areas of economic/political reform.

  15. I find the Swales approach to be more conceptually difficult. A basic income is not expensive if it is low and it replaces existing social security and the existing tax free threshold. A basic income also covers those on benefits other than the unemployed. I’m not dismissing your idea but I do see some difficulties in selling it.

  16. Terje, I came at this through game theory, so once I had the first insights (about phase changes, as it happens), it sort of fell into place for me; for someone like me, the trail was obvious. I had it all worked out before I ever found out about Kim Swales’s work along the same lines, which he reached by a different path.

    But I’m not typical; I have the remains of a very good mathematical background, though I do say so myself. So I tried to write it up with analogies, taking the mathematics out. Perhaps you might like to look at my first published article on it? Either that makes it clear, or – more likely – you will find it puzzling, and your feedback would help me make it clearer without turning out something untrue instead. Yes, it’s a hard sell if it shoots over people’s heads – but I don’t want to retreat into sneering at them for not getting it.

    For what it’s worth, the way Basic Income hits real life problems is that there’s a hard sell getting people to allow wages to fall, first in legislation and later from “stickiness” as they bargain, and even a “small” amount of Basic Income per head multiplies out to something huge over the whole labour pool. Then people would treat their new income as “free” for a while until they understood it, and so they wouldn’t appreciate it properly at first and there would be an inflationary pressure while they were still spending all of it… It would be a long haul and a rocky ride until things settled, and there might not be the political will and/or economic resources to get to the other side. What I would actually want to see is a start the Swales way, a later shift of weight to a Basic Income, then take the government out of the loop by turning supplying that into an individually endowed fund system – but as I said, that’s another story.

  17. “A basic income also covers those on benefits other than the unemployed” – Terje, in case you missed it, the Swales approach also benefits those people indirectly if it is implemented with anonymous transferrable vouchers, since they can be “cashed in” even without having a job; it’s how you transition towards a Basic Income system. (You can bet bureaucrats would be unable to see any alternative to an id based system they would be needed to run, though.)

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s