There is currently a push on in the United States to
increase (frequently the word ‘double’ is used) the minimum wage. This is part
of a push for greater equality in American, and part of efforts to ensure that
the working class benefits from America’s consistently strong growth.
Unfortunately, despite these good intentions, it’s quite possible that a move
in this direction will actually be catastrophic for the working class.
One of the most basic ideas in economics is the price
elasticity of demand. If price goes up, demand goes down. There are very few
exceptions to this rule. Within certain boundaries, demand for luxury goods can
increase with price. Bling H2O is one example. There is also a
theoretical class of goods called Giffin goods that have a positive price elasticity
of demand. Some economists speculate that potatoes after the Irish potato
famine displayed this property. Price rises led people to believe that potatoes
were about to become scarce, so they bought more potatoes. More recently, there
is the case of toll-ways in America that increase in price the more people use
them. There is a natural tipping point where the price is so high that people
think the toll-way itself must be congested and instead opt for the regular
highway, thereby ensuring an efficient distribution of cars across the two
services. For the most part though, if some increases in price, consumers shift
to alternatives (such as from quinoa to rice) and otherwise reduce their demand
for that category of goods.
The same applies to labour. When the price of labour
increases, the demand for labour goes down as firms try to use relatively
cheaper capital inputs (machines) instead. This is quite relevant in the
context of minimum wage workers in the developed world where, for example,
automated tellers are replacing check-out operators because they are cheaper.
Let’s look at this graphically. Below is a stylised (i.e. very simplified) rendition
of the entire labour market (figure 1). Where the demand and supply curves
interact we find the equilibrium price (average wage) and quantity (total
labour employed). When the wage increases in figure 2, we find that the
quantity of labour demanded falls (it's not worth paying that much) and unemployment emerges. This is what we see empirically when we
compare changes in the average wage to changes in total employment. A running
example is given by the flying geese phenomenon in East Asia (or even just in
China), where firms relocate their factories to new regions and new countries
as the local wage level rises (though workers left behind typically find new, better jobs because they have high human capital, which is precisely why their wages rose).
The price elasticity of demand theory has been borne out
countless times for the class of goods known as ‘labour’.[1] However, the empirical
evidence for employment effects from changes in the minimum wage has
historically been more mixed. For a long time, studies only found very small
impacts on employment levels when the minimum wage was increased.[2]
It is these studies that the people most keen to
dramatically increase the minimum wage tend to cite. The volume of these
studies means that at the very least, advocates can say that ‘the empirical
evidence is inconclusive’.
I find that quite a hard pill to swallow because in the
academy, this kind of thinking has been destroyed in the last 10 years. The key
problem with the literature that finds only small impacts from changes in the
minimum wage is that it inevitably focusses on average employment rather than the employment of people on the minimum wage. But changes
to the minimum wage will only affect people earning near the minimum wage and
not others. Moreover, as the vast portion of labourers earn in excess of the
minimum wage, it is unsurprising that tests of the impact of the minimum wage
on average employment find only small effects. (Another problem is that these studies often fail to control for general growth in the economy, which drives up all wages and demand for minimum wage labour, but let's not worry about that here).
This analysis is perhaps easiest to understand
graphically. Consider the following stylised depiction
of the distribution of workers by wage. Most people earn close to the average
way, which is near the hump in the middle. Some people are super rich and some
quite poor. Note that the people at the bottom of the distribution are probably
not earning a wage at all but are actually on welfare. When people lose jobs
they go onto welfare.
What happens when we introduce a minimum wage? The
distribution of workers near the minimum wage changes. Some lose their jobs as
their labour becomes too expensive and they are replaced by automated tellers or
teenagers who aren’t entitled to the minimum wage or the business decides not
to open on Sundays. These people either become unemployed or find black market
work at less than the regulated wage rate. Those who retain their jobs earn
more money. Critically, the shaded area under the curve is less than the area
above because of substitution effects.
The story depicted above has been borne out by empirical
work. Estimates of the minimum-wage elasticity of demand for minimum wage
labour can go as high as 0.7.[3]
Now that we have this understanding, let’s cut to the
policy chase. If you increase the minimum wage you will put a substantial
number of people out of a job, especially where substitutes in the form of
machines or foreign labour are available. If there is no welfare safety net
this will be a disaster for them. They will likely become homeless and
dependent on charity. As the United States has very little in the way of welfare,
this outcome is likely to be widespread following a doubling of the minimum
wage.
However, the current situation in the US for minimum wage
workers is repugnant. Minimum wage workers often work multiple jobs and a lot
of overtime just to get by, don’t time have to parent their children (leading
to entrenched disadvantage and increased crime), are very vulnerable to health and
crisis shocks (like deaths in the family), and frequently rely on the earned
income tax credit to make ends meet (which is not the purpose of that policy). They have very little potential to escape
this cycle because their wages do not permit them to earn enough to train
themselves for better work. The kinds of jobs that they are qualified for are
becoming increasingly rare owing to automation and offshoring, so there is
downward pressure on their wage which means this situation is unlikely to
change on its own. They are, it is quite reasonable to say, wage slaves. In this context, something must be done, and increasing the
minimum wage seems reasonable. I hope I’ve shown though that this might not be the
case.
It is critical to understand that the minimum wage is a
welfare policy and must be delivered as part of a welfare system. You can have
a high minimum wage if the people made redundant are caught by the safety net
and ushered back into the workforce at a later date. This can be achieved
(though admittedly the quality of research in this area is shallow and the
number of success stories few) through educational interventions, job-training
programs, income contingent loans to relocate to where there is work, and other
policies. Importantly, it can also be achieved by ensuring that people like
single mothers from low socio-economic backgrounds, the mentally ill, and the
frail i.e. people who we may not even want working full time for a variety of
reasons, are the ones that bear the brunt of redundancies and earn welfare
rather than income while young, unattached individuals or members of potentially dual income households looking to build a life for their families without compromising their ability to oversee their children get the jobs and are
thus granted the long-term mobility that provides. In this way the more
pernicious effects of both unemployment and welfare can be mitigated
simultaneously. There is so much complexity here that I can't go into. However, I want to stress that I am a big fan of both welfare and the minimum wage (and taxation and progressive economic and social policy in general), but the minimum wage on its own has a lot of potential to be a bad policy.
Now getting a single bill passed in the US is an
incredibly difficult undertaking, so it is understandable that advocates for
society’s marginalised there want to take things one step at a time. But
focusing exclusively on the minimum wage has the potential for disaster. At
the very least, advocates should consider that maybe the people opposed aren’t
just bastards and fat cats of industry, but also people who just want a bit
more nuance in how problems of this kind of complexity are approached in policy
and public debate.
P.S.
There is a relevant area of research here relating to the work of Card and
Krueger on monopsony purchasers of minimum wage labour in the US, notably fast food chains. I did not
include it because the piece was already very long and because that literature
remains contested. In any case, that literature may well be relevant to certain
states and districts, where a higher minimum wage can be rolled out. It needs
to be investigated case by case. Meanwhile, pushes for the minimum wage seem to
be targeted principally at the federal level. The original book is Card and
Krueger (1995) ‘Myth and Measurement: the new economics of the minimum wage’.
For a more thorough treatment than what I did here, have a look at the Lewis reference in footnote #3.
[1]
For example, Lewis & Seltzer (1996) “Labour Demand” in Norris & Wooden “The Changing Australian Labour Market”,
AGPS Canberra; or Lewis & MacDonald “The Elasticity of Demand for Labour”, Economic Record vol. 78 No. 240 pp. 18–30.
[2]
For a review of these studies, see Seltzer (1997) “An evaluation of international
evidence on the employment effects of minimum wage legislation” in Australian Economic Review, vol. 30 No.
2 pp. 208–14
[3] See
Lewis (2006) “Minimum wages and employment”, Australian Fairpay Commission
Research Report no. 1/06
I don't understand why you say that helping people make ends meet is not the purpose of the EITC. Could you explain?
ReplyDeleteThe EITC is meant to incentivise people on welfare to work. In a sense, it reverses the traditional problem of welfare, which is that as you earn money you are entitled to fewer benefits, so you are disincentivised to work. With the EITC, there is an incentive to work while on welfare, which lubricates the transition back to the workforce from welfare.
ReplyDeleteIf someone *relies* on the EITC to make ends meet it means that the economy is f**ked. That is a situation where someone has secured work but is nonetheless unable to make enough money to live. This suggests that there is a substantial divergence between the cost of living and the market wage, which seems to me to suggest policy failure somewhere.
As far as I can tell from my admittedly limited reading, the EITC was designed to keep some welfare recipients, notably single parents, working at least 10-20 hours per week so that their children would view work as important. The purpose of the policy isn't to replace welfare, it is to send cultural messages and help people transition back to work (for example, people start as casuals, claim the EITC, then pass probation and move to full time, thereby exiting the EITC). In my opinion, the way it is currently used, where people often work full time hours and still don't breach the EITC threshold, represents a failure of social policy in American on a deep and broad level. Notably, human capital production must be catastrophically bad or else people would command higher wages even if they were relatively low-skilled. That said, I don't know how widespread that phenomenon is.
I understand the countervailing incentives you describe between welfare benefits and the EITC. That makes sense to me and sounds like a good thing. But, I still don't see why it is bad that some people have to have their wages subsidized to make a decent living. I'm not sure why you infer policy failure here. Could it be a market failure in a sense? This is a stretch from the textbook types of market failure, but maybe inequality can be construed as a market failure, and so the question is what policies alleviate it. It seems to me you might be calling "policy failure" while dismissing one of the better policies suited for the purpose.
ReplyDeleteYou seem to suggest that lack of human capital production is a fundamental problem causing inequality. So, you aren't saying the EITC is a problematic policy in itself, just that, as used in the US, the EITC is indicative of other problems?
I like the EITC; it's certainly not a problematic policy. I also don't think a wage subsidy is a bad thing, especially if it is done explicitly to improve equality. Inequality isn't a market failure because equality is not something that markets do. They merely efficiently allocate resources to their most productive uses. However, while markets are good for efficiency they aren't good for equity, so you should certainly buttress them if you care about equity. The EITC is one way of doing that.
ReplyDeleteI think I'm probably just not being up front about my priors. I don't think work is strictly a good thing, so I am not inclined to force people to work terrible jobs full time just to grant them a wage subsidy. I would honestly rather have a higher minimum wage, which gives people an incentive other than starvation to take bad jobs, and then simply provide welfare transfers to the people who are incapable of holding such jobs because of mental problems, catastrophically low IQ, histories of abuse or general dysfunction. I can see the merits in having such people occupied regardless, so I would have things built into welfare, like the EITC, to have people work 15+ hours per week, and I would have volunteer for the dole (rather than work for the dole) type programs, so that people contributed to their communities for their welfare rather than just digging holes and filling them again.
My problem isn't really with the EITC; it's with the attitude that work is fundamentally good. Conservative puritans, including here in Australia, seem to think that if the disadvantage would just work like slaves they would inevitably end up on the straight and narrow. I don't agree with very much at all.
We can avoid forcing people to choose between terrible jobs and starvation by providing welfare entitlements that are not means-tested. As non-means-tested benefits become more generous, the reservation wages of the low-income workforce will rise. If people can live decently without working, and, as a result, they are empowered to demand higher wages, then the minimum wage seems pointless. If we want people on welfare to work more or if there is a shortage of low-wage labor, then raise the EITC. In this scenario, the minimum wage is unnecessary. What problems do you see in this?
ReplyDeleteSeems pretty good me. I am increasingly a fan of the basic income payment. The only problem I can foresee is that a lot of people like myself would move to remote coastal locations and live leisured lives rather than being financially productive members of society, or at least retire early. This is not necessarily a problem though.
DeleteTo be honest, it's hard for me to comment much on the appropriate way to manage the US situation because it is so different to the environment that I most familiar with, namely Australia. This article was merely meant to outline some of the contours of the minimum wage debate because I find so much of the commentary on it clumsy, included, oddly enough, Russ Roberts on Econ Talk.
Thanks for your continued comments.
Another problem is of course the political economy of increasing welfare in America, but that's I guess precisely where the political discussion is at in the United States. Your suggestion is leagues more sound that doubling the minimum wage.
Delete