The living wage again: Considering policy choices and trade-offs

I see a new living wage has been released (Via Kiwiblog).  We have posted about this recently, briefly noting some areas that need to be considered when thinking about this concept.

To help think about policy choices more broadly, Richard Meadows asked a few economists about making the minimum wage equivalent to the living wage (not what the living wage proponents are suggesting – but it is something people keep suggesting) and removing Working For Families.  The results of his conversations were here.

It was a good article, although I would note that I didn’t intend to come out quite as harshly against a tax-free threshold as I did – conceptually I see issues, but I want to leave my mind open until there is modelling work done to articulate the impacts and any potential trade-offs.  I did enjoy seeing the conclusions:

Crampton says he simply can’t say whether it’s better to take money from childless people – both rich and poor – and hand it over to low to middle income parents.  Economists can talk about the trade-offs involved and the likely effects of the policy, but moving beyond that becomes a value judgement, he says.

Hazledine is prepared to take a stand. ”I have long believed that excessive population growth is one of the major contributors to the problems of the modern world,” he says. ”So I am opposed to any pro-natalist ‘buying babies’ policies in any country, including our own.”

Nolan has read literature saying we should have higher birth rates – and other material saying just the opposite. ”Ultimately, this is an open question, and one that I don’t think is particularly useful for policy.” Instead, he says the best way to think about it is to consider people’s needs – and the trade-offs from the policies which are trying to address them.

I’ve asked Richard if I could put up the notes I scrapped together when discussing his piece – and he said that is cool.  As a result, my views and some of the questions that were asked, can be found below the flap.

This was written as a word document, and I have just pasted it across.  It hasn’t been edited into a blog form, hope people can still make use of it.  Would be happy to discuss the ideas in the comments, I’m a touch busy so it may take me a while to reply though 🙂

Note:  Eric does his write up here!

Thoughts on the minimum wage, tax free threshold, simulations

As a broad point for all of this, the key interest is in medium/long run effects from tax and benefit policy – not short run effects.  The impact on short-term aggregate demand etc etc is a different issue, and one the RBNZ will be trying to deal with directly.  This is structural policy, so government sets these with its eye on what happens over a long period of time.

The very first point to note when thinking about the scenario is how many people actually fall into the group you are looking at.  Treasury recently looking into household level data for the minimum wage in order to understand what having the minimum wage at the “living wage” level would mean:

A key thing to keep in mind from the start is that the minimum wage is a price floor, and so it only matters in so far as it “binds”.  In the Treasury analysis (an analysis I can give you a bit more technical detail about on request – and which Treasury themselves are more than happy to discuss), they note that:

  1. Only 15% of the families earning below the living wage are in the category “2 Adults with children” category.
  2. Retail related industries are the key employers in this category, and as a result this is the area where we would expect to see prices rise/output and investment fall.
  3. Figure 4 in the report notes that most of the “additional income” paid from employers will go to the government in the case that this is introduced – through taxes and abatement of benefits.
  4. A “living wage” as a minimum wage would put us well above the distribution of minimum wage/average wage paid in OECD countries.

These points will all be useful when it comes to discussing the impact of the changes in the scenarios.  [Note:  If you do have a little bit of time, I would suggest skimming the Treasury report – as it adds some additional points about mobility and where the burden and benefit fall].

When it comes to thinking about need these questions can be difficult – which is why it is very much talking about trade-offs from changes in policy settings, rather than getting ourselves bogged down in saying which is best.  A few posts I’ve done looking at the discussion of need (which might be useful) are here:

Starting questions

To what extent do childless people subsidise families under the current model?:  Households with children are certainly subsidised by government – but to a lesser degree than many other countries.  Some key points to consider are:

  1. Our tax system is based on an individual’s income.  Our benefit system focuses on household/family income.  In this environment:
    1. WFF is targeted at working families with children.
    2. The benefit system also pays more to solo-parents than to non-parents.
    3. Government expenditure on facilities tends to fall more heavily on families eg education.
    4. However, unlike many other countries we do not allow “income splitting” for families – which would be a form of tax benefit for co-habitation.

Is the subsidisation of families considered a “good” thing for the economy as a whole?:  In the broadest terms there isn’t a consensus for this question – I have read literature saying we ‘should’ have higher birth rates, and read literature saying we ‘should’ have lower birth rates.  Ultimately, this is an open question – and one that I don’t think is particularly useful for policy.

The best way to think about it is to think of ‘need’, and the trade-off’s associated with policies that try to address ‘need’.  In that context we can think of households in the following way:  A household is an economic and social unit, where individuals choose to band together for some reason.  One of the key benefits of joining together is “economies of scale” whereby a household can meet its needs/wants more cheaply when the individuals specialise.  I briefly summarise the idea on point four here.

Government subsidises can be seen in this context.  Rather than thinking there is something good about having more kids it is more about two things:

  1. An implicit preference by society that they prefer households with children to households without children
  2. A view that parents may underinvest in children, and therefore it is a good idea for society to subsidise/invest in them directly.

As a result, this is the context we need to think of household subsides through.

Are there unintended consequences?:  There are two key consequences of the current system:

  1. 1.       Subsidising children will lead to people having more children
  2. 2.       The way we subsidise promotes labour market entry and greater labour supply

Now these are not unintended, they were there by design.  However, I am not quite sure whether we as a society have thought them through clearly.  Especially the labour supply argument – there is a “cost” to supply labour (something we don’t take into account if we only look at the fact GDP is higher), and we are excluding households which are unable/unwilling to work.

Scrap WFF, increase MW

This looks like a change in focus of the tax-benefit system away from supporting households, and towards supporting single people.  However, as you go on to note later we can’t really conclude this until we know how the government is “spending” the money they save on WFF.

The WFF cuts and the MW issues are quite separate.  I’ll focus on MW here, and touch on WFF in the following section where you say how you are spending the money.

What impact might hiking the minimum wage to $18.40 have on unemployment?:  At this level, the minimum wage would cause unemployment.

There are two key issues here, the level of the minimum wage (and therefore the amount it is binding) and short run vs long run effects.

The debate in the US is about lifting the minimum wage, but to a level that is still relatively low (compared to average wages).  In that context, there is a lot of good evidence that there is little short run impact on employment or hours, and many economists support the idea of lifting that minimum wage – given the view that employees don’t have market power:

However, the living wage is very high relative to the average wage (as noted in the Treasury piece).  As a result, it is both binding for a lot more people, and is also considerably more costly for firms.  This implies that we can expect lower employment in the short run.

The big impact isn’t in the short run though, and this isn’t what concerns economists the most.  It is the long run impact, through disemployment (firms deciding not to invest by hiring) and investment (firms deciding not to invest).  A good write up about that is here:

Both the short run and long run effects will fall disproportionately on certain groups – namely young people (and the retired) who are looking to earn income in the retail/accommodation sectors.  By knocking the young out of the labour market we limit there opportunities and choices with regards to building up human capital.  Some may quote this excellent DOL report by Hyslop and Stillman to say it works the other way:

Given that report found people were leaving the labour market to study.  But this doesn’t answer the question of whether people are investing in the right skills – remember, people are deciding to study because they have fewer options, this is hardly conducive to making the best life choices.

Update:  realised I missed an important point.  The higher the minimum wage, the higher the cost for a firm to hire someone – as a result, employees that can’t “create enough value” will get excluded from the labour market.  This implies that the higher a minimum wage is, the more people get excluded from the labour force because, for some reason, they are not able to do enough (eg the disabled).

This is one of my main issues with making the minimum wage too high, it leads to people with mental and physical disabilities being excluded from the labour force – as the jobs they would have worked either get canned or replaced by capital expenditure.

As a result, such an indirect means of targeting can exacerbate other social injustices.

Would the costs of goods and services rise?:  This is a good question, the answer is yes for direction – but the magnitude is a harder one.  We know the cost pressures will fall in the retail sector, given global competition we will see prices rise more in the non-tradable part of retail (haircuts, McDonalds) than in the tradable part (books and DVD’s).

We can dig a bit deeper though.  If we are arguing that the minimum wage increase is good because it is a transfer from capital owners to labour, and that there will be very little impact on employment as the firms are monopsony’s ( ) then we can add some more logic here.  A monopsony is likely to also have monopoly power in goods markets, implying that they can pass on cost pressures – and so prices will rise.

As a result, if we are arguing that the high minimum wage will be an effective solution with very little cost to employment we are ALSO arguing that much of the cost pressure will be passed on to consumers!!!

Tax free threshold

I think I can sort of answer all the questions in this section by just chatting about tax free thresholds.  It is a hard issue I would like to see modelling work on before making a firm conclusion!

However, conceptually we could make the argument that a tax free thresholds is probably one of the worst possible tax cuts that could ever be put in place (for meeting any given equity/efficiency goals).  They exist overseas as they are really easy to market as an election bribe, but they have many poor properties:

  1. It is essentially a lump sum transfer to anyone earning above the threshold – this makes it very expensive without changing marginal tax rates for a lot of people.  As a result, it will not have much of a “labour supply” effect.
  2. If we are aiming for the same “size of government”, then a tax free threshold implies we need higher marginal tax rates for the same “average tax rate” (which is the size of government).  So it reduces labour supply and investment incentives!
  3. There is a common argument from the 1960/70s that it reduces “churn” in the tax system.  However, with computerised accounts and the such the churn argument is close to inconsequential nowadays for a tax free threshold.

When taking into account the change in WFF, we can also say:

  1. WFF did increase labour supply of second earners – part time employment and employment rates rose considerably in WFF’s wake.  As a result, removing this and introducing a policy that is likely to reduce labour supply incentives will have the net impact of reducing labour supply, and reducing GDP.  [Removed as the claim is unsubstantiated – and not even something I agree with in the way I framed it 😛 ].  Given the strong lift in the participation rate post WFF, it may be the case that WFF was behind this lift (through solo parent entering the labour force).  If this was the case, removing this, especially when combined with a policy that is likely to reduce labour supply incentives may have the net impact of reducing labour supply, and reducing GDP.
  2. If we, as a society, decided that there was a “need” that WFF was filling then we are swapping a policy that targets need with an untargeted policy, so in social terms this is likely to be a harmful change.
11 replies
  1. Eric Crampton
    Eric Crampton says:

    Where’s the evidence on WFF and second earners? Theory would have said that EMTR on second earners went from (whatever second earner’s MTR would have been) to (that plus WFF clawback); this should have reduced second-earner labour supply. Have we data showing otherwise?

    • Matt Nolan
      Matt Nolan says:

      Ahhh that was lazy of me at the end – I should update that to be more conditional.

      I’m not aware of any actual estimation work post-reform – only simulation work prior to the reform. And that suggested that the positive employment effect of the in work tax credit would be small, and that there would be a disincentive effect from abatement.

      It was one of those situations where I wanted to say “given the strong lift in the participation rate post WFF, it may be the case that WFF was behind this lift. If this was the case, removing this, especially when combined with a policy that is likely to reduce labour supply incentives may have the net impact of reducing labour supply, and reducing GDP”.

      • Eric Crampton
        Eric Crampton says:

        I would totally expect increased LFPR among single-parent families. Then household income equals that earner’s wage income; the wage subsidy draws those who hadn’t been in the labour force into the market. Clawback rates on WFF might stymie progression a bit once they’re in the market, but the big effect would be getting those people into the labour force.

        You’d need individual-level data to do this up properly – it’d be a big job. I don’t think we can just look at aggregate LFPR, which dipped a bit among women post WFF then came back. We really need to know the by-cohort EMTRs, whether they were in-work or not prior to WFF, kids’ ages, partner’s incomes… no way I can do it from my desktop. Somebody would have to do the special pleading to go sit at a StatsNZ data centre and start working it out.

  2. Steve
    Steve says:

    To what extent is working for families a subsidy for employers? That is because it is tied to actually working employers can pay lower than an equilibrium wage because those families are receiving a subsidy to the parents’ wages. By tying working for families to actually working (rather than just children) it subsidises the wages of working parents indirectly to companies. Employers can pay working parents less than equilibrium (assuming equilibrium is above the minimum wage which it will be for many people on WFF) and this could also push childless people out of work as they are relatively more expensive to hire since they no longer receive this subsidy. Swapping WFF for a higher minimum wage or just a child subsidy would take away this subsidy to employers.

    But what about a basic income policy as a better alternative? I’m surprised none of the economists proposed this in the discussion. It would mean the subsidised income isn’t tied to working, so doesn’t end up in the pockets of companies. It also reduces inequality without discriminating between families and the childless and doesn’t increase unemployment. Some people may choose not to work under this policy, but that isn’t necessarily a bad thing. People have that choice. There is also an argument that it allows entrepreneurs to pursue innovation rather than otherwise feeling required to work in order to support themselves. It could even be a varied amount depending on whether our collective value judgement is that those with children should get a greater basic income.

    • Matt Nolan
      Matt Nolan says:

      “To what extent is working for families a subsidy for employers?”

      This depends on the elasticity of demand and supply of labour right – it is a question of tax incidence. A very important point when thinking about designing the scheme, and how it meets need!

      One thing I’d keep in mind is that employers also have to pay single people the same rate as those in families – and WFF creates a wedge between the return to labour for these two groups. As a result, those with children will drive down the net wage received by people without children. As the questions were focused on this area that is another point.

      “But what about a basic income policy as a better alternative? I’m surprised none of the economists proposed this in the discussion”

      We were given scenarios and asked to comment on them – we didn’t really have space after all that to suggest the article should be on something else! Saying that I’m a big fan of minimum income schemes (, and Eric and I have both written on them. I like Eric’s post discussing the trade-offs we need to think about with such schemes:

      I would be a bit careful with your conclusions about it though – you are definitely on the ball with a lot of the areas of interest, but many of these things are empirical questions that we can measure and then discuss. Furthermore, it is definitely not clear that such a scheme would reduce income inequality!

      One thing I would note with minimum income schemes is that New Zealand has a minimum income scheme – but we also tie it to work testing. In reality, supporters of the GMI are saying two things:

      1) Remove work testing

      2) Remove targeting

      They both involve value judgments – and in fact the second one is very harsh. Oft time too many things are chucked in the “GMI” basket – there is a lot of debate to be had around the trade-offs that exist for different specifications of a GMI!

      • Steve
        Steve says:

        Thanks. Yes these really are empirical questions. I am assuming that demand for unskilled labour is relatively inelastic in that they are jobs that cannot easily be replaced by capital goods in production. With this in mind, the incidence of the subsidy could be one to employers as I suggested. But yes, it is an empirical question. I suppose it also depends to what extent supply of unskilled labour is inelastic. And this too will be related to their alternatives for income. With no support for those who choose not to work, supply of their labour is also inelastic sharing the incidence between employers and workers. But this supply would be much more elastic if they had a basic income not tied to working. As a result, of the basic income thing, it might also remove the need for the minimum wage. Sorry I don’t mean for these to be conclusions. They are more hypotheses where I am interested in your opinion. They would be interesting things to study empirically.

        the wage thing is also true, I had forgotten that employers cannot discriminate between the childless and parents. So the result is not so much that people without children are unemployed, but that they might choose to exit the workforce, except as I describe above their supply of labour is probably inelastic, resulting in unemployment as I originally describe.

        I suppose the GMI inequality thing might not be as I predict. But it would reduce poverty (by either measure) which is my main concern with increasing inequality. So I am also in favour of it.

        I have no problem with targeting but it is how that targeting is implemented. This is also the issue with WFF in that as parents earn more, their WFF is reduced very quickly meaning they have a very high effective marginal rate of tax. If we choose targeting as our collective value judgement it needs to be implemented in a way that the effective marginal rate of tax is not so high as to discourage earning more. I do have a problem with tying it to working because I hypothesised that this means at least some (if not most) of the incidence goes to employers.

        • Matt Nolan
          Matt Nolan says:

          Hi Steve,

          Good comments. I agree with you that removing the requirement for job search may well remove the need for a minimum wage. In fact, the very nature of changing opportunity by doing this could lead to a big change in the relative wage rates for different forms of work – which would make such a question very difficult to think through. However, conceptually it fits our idea of “opportunity” fairly well, which is nice.

          There are a lot of details though, and it would take a team of people far smarter than me to make headway on them! Hopefully that team turns up soon 🙂

Trackbacks & Pingbacks

  1. […] it into his excellent story, here. I agree with everything Matt Nolan said there too.[Update: Nolan's extended comments are here. And I'm really pleased to see that Richard's linked the more extensive commentary from both of […]

Comments are closed.