Minimum wage and tax on low income earners

By Matt Nolan 12/04/2013

What would you say if I told you I was currently in control of a hypothetical little country in my head – other than accusing me of of having some form of psychosis.

In this little country there is a group of low income earners, who have a reservation wage (the hourly satisfaction from growing vegetables in their own garden) of $6hr.  There are also a bunch of high income earners that do what they do, and who the low income earners can’t replicate.  The kicker is that, the work these low income earners are able to do is only available from a cheeky monopsony employer!  This employer is able to extract $15 per hour of value from each of these workers, and since it has all the market power it only pays them $6 an hour.

Ok, now I’m a benevolent government in my mind – and I’ve decided this isn’t fair.  I’m trying to decide whether I should turn up an demand a $10 minimum wage, or whether I should put a 40% tax on just low income earners and then throw the money back at them.

Now you might think that what I just wrote was a typo.  Surely I meant “cut” taxes on low income employees – after all, higher taxes mean less income!

Some keen eyed may have noted that I said “throw the money back at them”.  They may say “what’s the point, you are taking money off them, having to pay for ‘churn’, and then giving them less back”!  Now for kicks, assume there is no churn in my tax system – it is in my head after all.

The majority of people at this point think that the tax scheme is just me being weird, and that it makes no difference.  However, the minimum wage would increase the wage, so lets do that.

But, what would you think if I told you that, given all these assumption, they actually have the same result!


Many people would say that the tax is 0.4x$6=$2.40 per hour and the transfer back is $2.40 per hour so it cancels out.  But this involves not considering the incidence of tax.  Our low income employees have NO bargaining power, they are paid the least possible by this monopsony employer.  However, they will not work for less that $6 an hour that is their reservation level.  They still need to be paid this NET of tax in order to supply their labour.

As a result, the firm needs to offer to pay a GROSS wage of $10 an hour – $4 is paid in tax, which is then given to the low income employee regardless of whether they work!  So by taxing the employee and then giving them the tax money back irrespective of work, there effective wage goes up to $10 per hour!

But this isn’t realistic!

Indeed, it isn’t in the slightest.  However, here we are using the assumptions that are common regarding why a higher “minimum wage” is good and showing that those same assumptions also suggest that we could tax low income earners more and transfer the money back to them – and they end up with higher after tax income!

Now the devil is in the details.

  1. If the tax money is only returned if the employee is working, then we don’t get this result.
  2. It is unlikely (in either case) that there would be “no employment effect” – the firm will still not hire, or employee hours, of someone they would have otherwise.  Note that if we refuse this assumption in the case of increasing taxes only on low income earners we are ALSO refusing it for the case of minimum wages!
  3. We don’t really have monopsony employers anymore … but again this is a common assumption for the minimum wage argument.
  4. Fundamentally, both the quantities of supply and demand will “respond” somewhat to changes in price (the wage), and as a result the incidence of the tax itself is not so one sided.  However, if the transfer is without churn (and churn is pretty small nowadays) we are just taking that tax chunk that is ripped out of surplus and giving it to employees … just like we stated!  We just have the additional people who lose out from not being employed and the associated dead-weight loss.
  5. If we transfer the income to the low income group who are still employed and those who are not employed (so assuming lower employment occurs), then the difference between the tax case and the minimum wage case is that those people who get left out of work are better off in the tax case, to the cost of those who manage to get work (who are still better off in the no policy case in this example).
  6. The real dynamics, and what we can really observe and respond to, are entirely different to this – this is a thought experiment that isolates tendencies stemming from these policies.
  7. If the industry is a large part of the economy, this business gets even more complicated ….

What I would note though is that, the less market power we assume our employees have, and the lower we expect the employment effects to be, the closer we get to this situation – the more of the incidence of taxation ON these people actually falls on their employer!

A lesson

There is something pretty thought provoking to come out of this example though.  We have discussed a case where taxing someone, and then giving them back the money made them better off in a redistributional sense – because prices changed.  Even though we “taxed the employee”, the incidence of tax fell on the employer, and so we transferred income from one to the other.

Many people will look at a situation and say “I want to help those people, let’s cut tax” – but we need to actually ask about the incidence of tax, and the response of government spending to the lower revenue, before we can articulate who wins and who loses.  There is FAR to much arbitrary moralising by people regarding tax rates, without any clear indication that they have thought about tax incidence when coming to their conclusion.  And if you don’t think about the incidence of tax, you are stating an opinion on tax based on no thought … I have to be blunt.

As a secondary result, hopefully by “reframing” the minimum wage argument using different words, people can look at the issue with more critical eyes – yes it should provide a transfer, but who the transfer is between isn’t always clear!  The reason some people prefer direct tax credits and direct income transfers isn’t because we are being pedantic – it is because the actual distributional consequences are clearer and more proven … it is because we want to ensure society’s wish to help people (if it has this desire) does just that, rather than using poorly designed policies that merely “show the intention” of helping.