A year ago, I brought attention to a Canadian Public Policy article on poverty and the minimum wage in Ontario. A notable finding of the study was that the overlap between those who earn minimum wage and those who are in poverty was surprisingly small, small enough to conclude – as I did – that "increasing the minimum wage is only slightly more effective as an anti-poverty measure as would be distributing money at random across households." A few months later, I came across a study that found much the same results in the United States.
I knew that policy analysts in the Quebec government were working on a similar project and had heard that they had found similar results. That study has finally been made public, and is one of a set of articles on the minimum wage. Even better, it would seem that Pierre Fortin has fleshed out and put into digital form the literature survey I blogged about over here (has it really been four years?!?).
But it's this article by Jean-François Mercier and Martine Poulin that has the numbers I'm going to talk about below. A very nice feature of the study is that it pays particular attention to those earning just above minimum wage.
Here is a table of some numbers taken from Tables 2 and 4 of their paper:
| Low-income families | Families above low-income threshold |
All families | |
|---|---|---|---|
| Total | 448 700 (13.1%) | 2 979 100 (86.9%) | 3 427 800 (100%) |
| In employment | 193 900 (8.4%) | 2 105 300 (91.6%) | 2 299 200 (100%) |
| Minimum wage or less |
27 900 (13.5%) | 178 200 (86.5%) | 206 100 (100%) |
| $7.61/hr – $8.99/hr | 44 200 (18.2%) | 199 100 (81.8%) | 243 300 (100%) |
| $9.00/hr – $9.99/hr | 34 100 (17.3%) | 163 400 (82.7%) | 197 600 (100%) |
| Less than $10/hr | 106 200 (16.4%) | 540 700 (83.6%) | 646 900 (100%) |
As was the case in Ontario, the proportion of minimum-wage earners in poverty (13.5%) is only slightly greater than the proportion of those in the general population who are in poverty (13.1%).
It is at this point that someone invariably claims that those in poverty are clustered just above the minimum wage, and that these people are the real targets of the measure. Well, not so much, it turns out. The incidence of poverty in this group is higher than in the general population, but it never goes above 20%. Given this weak relationship, it's hardly surprising that their policy simulation exercise found that increasing the minimum wage had only marginal effects on inequality.
Moreover – as the authors note – this marginal effect is the best-case scenario, because it is based on the assumption that increasing the minimum wage has no effect on employment. Although the effects on employment are likely to be small, they won't be zero, and – as Pierre Fortin notes in his literature survey – estimates from Canada are generally larger than in the US. A study prepared by some of my colleagues for the Quebec government puts it this way (my translation):
An increase in the minimum wage can induce firms to reduce employment in the medium to long term. The literature from Canada is unequivocal on this point. Studies by Gunderson 92005), Baker, Benjamin and Stanger (1999), Benjamin (1996), Sussman and Tabi (2004) and Statistics Canada (2005) show that workers who are most vulnerable to these effects are those with the lowest levels of education, who have been in the labour market for the shortest period of time, and who have the least seniority at their workplaces.
Studies focussing specifically on social assistance in Quebec also show that laboour market integration is more difficult at high levels of the minimum wage (Duclos et al (1996), Duclos et al (1999), Fortin et al (1999), Fortin et al (2002), Lacroix (2000)). For example, Côté (2006) finds that regardless of the type of household, increasing the minimum wage has the effect of increasing the duration of spells on social assistance. In fact, the only groups that are not negatively affected are single men over 45 and single women under 25. The increased difficulty in finding employment and leaving social assistance is likely due, ceterus paribus, to reduced labour demand on the part of firms.
You can check out the references in the original report. You will note that the minimum wage file is tangential to its main focus, which I'll be revisiting in the near future.
The employment effects of the minimum wage might be safely ignored if the gains from doing so – in terms of reduced poverty and inequality – were "large enough". But available evidence makes it clear that these gains are far too small to justify ignoring employment effects. Moreover, the data suggest that marginalised workers are most likely to pay the price of those job losses.
I've said it before, and I'll say it again: as anti-poverty measures go, increasing the minimum wage is pointless at best.
Worthwhile Canadian Initiative 
That is eye-opening. I have always been very suspicious of the “minimum wage as redistribution” argument. My suspicion has now become quite firm.
I’d be curious in knowing how many people who earn minimum wage are highschool or college/university students. Stating that the minimum wage is important because it helps students pay for rising tuition might be a better argument in favour of higher minimum wages, but curiously it’s not one I hear made too often.
I guess the biggest effect of the minimum wage on redistribution is on within-household distribution. Gives stronger bargaining position to MW recipients.
If MW recipients are the secondary earner, this gives them a stronger ‘outside option’–the threat of leaving is greater so they will have stronger bargaining within the household.
But for teens or young adults, it seems different. Do teens benefit from more bargaining power? For young adults, the idea of a threat point for exit might go the other way. That is, parents might want them to exit and a higher MW makes it easier for parents to give their young adult child the boot.
Frances knows a lot about these kinds of household bargaining models. I wonder what she thinks? Have these models been extended to relationships between older teens and younger adults and their parents?
Mike: Many of the relevant 2005 numbers are here. Haven’t found an updated version but should be similar.
Kevin: I hadn’t thought of that. OTOH, it’s not a line of reasoning I’ve seen articulated by minimum wage advocates.
Makes sense. I’m curious though…is there any research suggesting that there’s an inflection point where a lower minimum wage will reduce wages for those in poverty by more than it will increase employment? (I think that was poorly worded. Basically, from a theoretical standpoint, are we supposed to be better off by having no minimum wage at all, or does there exist an ideal point where the benefits outweigh the costs?)
I also wonder whether minimum wage doesn’t have an impact much higher up the income ladder than just the “under 10” crowd. It seems at least plausible that minimum wage provides a reference point that has an impact on negotiations for higher wage jobs.
Perhaps without minimum wage legislation, labour’s share of national income would decline somewhat? If so, a higher percentage of families would end up as “low-income”? (I’m assuming that returns on capital tend to accrue to the wealthy, while returns on labour tend to be spread more evenly, and firms tend to maintain a premium over the minimum wage, while a modest minimum wage has no appreciable effect on employment.) Just a thought.
Why so reactionary in opposing higher wages for crappy service sector jobs? It is still a good thing to have minimum standards in the workplace. And if we drove minimum wages up to where they should be — like $15 an hour in the big Canadian cities — it would have a positive anti-poverty impact. Would that benefit teenagers living at home, too? Yes, but that is also a good thing (have you seen what it costs to go to university these days?)
That said, a much better way to raise wages in the low-income service sector would be for government to precipitate a mass unionization of the sector, and have a level playing field for companies, who could pass off any cost increases to consumers.
It is important to let the labour market do much of the heavy lifting in developing a more egalitarian society. It cannot all be income transfers to the poor (though a guaranteed income should be part of the package, too).
Minimum wages benefit the working poor. Stephen’s table indicates that a majority of low-income employed families have hourly wages under $10 (i.e. 106,200 out of 193,900).
By comparison, only one-quarter of employed families above the low-income threshold have wages under $10 (i.e. 540,700 out of 2,105,300). So, raising the minimum wage is significantly redistributive among households with employed members.
It obviously does not directly help jobless households (i.e. the majority of “low-income families” in Quebec). We clearly also need other policies, such as income-support programs, to assist those families. But that is not an argument against minimum wages.
Stephen commented, “It’s not a line of reasoning I’ve seen articulated by minimum wage advocates.” In fact, Andrew Jackson of the Canadian Labour Congress often argues that higher minimum wages increase the economic independence of women and young workers.
“Perhaps without minimum wage legislation, labour’s share of national income would decline somewhat?”
It could also increase somewhat, as lower-value tasks stay in Canada rather than being outsourced to India for $5-8/hr.
“Minimum wages benefit the working poor. Stephen’s table indicates that a majority of low-income employed families have hourly wages under $10 (i.e. 106,200 out of 193,900).”
It also shows that a majority of low-income families are not employed (43%), contrast that to families above the threshold (71%). What would those percentages be if we took Marc’s suggestion and raised the minimum wage to $15? Particularly when you consider that an increasing proportion of jobs do not require someone be in a particular physical location and thus can be outsourced to India or the Philippines.
“Minimum wages benefit the working poor.”
Which working poor? The ones before or after you raise the minimum wage? Low wage workers will on average be earning more after you raise the minimum wage than low wage workers before (No s***!). But some of the ones who were minimum wage workers before will be unemployed after. Minimum wage is class warfare of the poor against the very poor. Yet another example of trying to fit a price solution to an income problem. Bad policy.
Mike and K seem to have retreated from Stephen’s fuzzy claim that there is little relationship between low wages and low incomes. We are back to the classic debate about minimum wages and unemployment.
Plenty of empirical literature concludes that minimum wages do not increase unemployment (e.g. Card and Krueger). Minimum-wage jobs are concentrated in service industries like restaurants and retail stores that plainly do “require someone be in a particular physical location” and thus cannot be outsourced.
“Mike and K seem to have retreated from Stephen’s fuzzy claim that there is little relationship between low wages and low incomes.”
I think Stephen’s data stands for itself.
“Minimum-wage jobs are concentrated in service industries like restaurants and retail stores that plainly do “require someone be in a particular physical location” and thus cannot be outsourced.”
I would like to see data on that, but I wouldn’t be surprised if you are correct. Of course, this ignores the shift from labour -> capital. For instance considering what’s happened to checkout lanes in the last 5-10 years.
Fuzzy? I thought it was pretty clear.
And I had rather hoped that the empirical debate had progressed beyond “Card and Kreuger spoke the ultimate truth for all points in time and space”. Please take a look at Fortin’s literature survey. To put the numbers in context for the literature review, the average hourly wage in Quebec in 2006 was $18.87.
On the household bargaining thing:
I think of myself as being progressive, and I would really like to see more done for the working poor. Especially families. But I just can’t bring myself care that much about young people earning minimum wage. Especially young men. Is it really such a bad thing for young men to have little or no choice but to stay at home and/or stay in school (trade, uni, whatever) rather than earning lots of money and having disposable income?
During the heady days of the last boom here in AB wages went through the roof and labor was scarce. Nobody paid anything close to minimum wage. It was so bad that chains like McDonald’s where closing the restaurant at 10pm because they couldn’t find anyone to work an evening shift at the wages they were willing to pay. Why work in fast food when you could work in O&G or construction for 5 times more? As a result, lots of young people – notably young men – who would otherwise have been living in their parents basements and taking the bus were now living on their own and driving gigantic pick-up trucks (with the corresponding increase in insurance rates). What we noticed in our economically diverse neighborhood was this: During the summer months the demographic in question has that habit of getting drunk, issuing forth from their dwellings, and fighting with each other and/or their girlfriends (also drunk) in the streets at 2am. We would call the police at least once a week to come round them up and toss them in the drunk tank. One young man was even stabbed outside the building down the street.
Then the recession hit. Unemployment went up and wages on the low end dropped like a stone. The midnight drunken brawling stopped, and the past two summers have been totally peaceful. The fast food restaurants and convenience stores are again open 24hrs. I suspect it’s because all the young idiots sold their pick-ups, moved back-in with their parents where they belong, and hopefully went back to school.
All-in-all, I think it worked out OK.
🙂
Erin: “retail stores that plainly do ‘require someone be in a particular physical location'”
They can also go broke.
“Plenty of empirical literature concludes that minimum wages do not increase unemployment (e.g. Card and Krueger). Minimum-wage jobs are concentrated in service industries like restaurants and retail stores that plainly do “require someone be in a particular physical location” and thus cannot be outsourced.”
I seem to recall that some of that literature suggested that minimum wages didn’t affect employment, it just drove down working conditions (i.e., you don’t lose your job, your boss just works you harder and makes you pay for your uniform).
Responding to Kevin Milligan’s question –
Short answer: in theory, yes, minimum wages should increase children’s bargaining power and that might be expected to have an affect on household outcomes.
Long answer: is there much research on this area that would tell us how much this might matter? Not really. From a theoretical point of view, 3+ player games are hard.
WCI reader Guy Lacroix has just done a paper that looks at whether or not children are decision-makers within the household. He finds that families consisting of two adults+a late teens daughter have three decision makers, while families consisting of two adults+a late teens son have two decision-makers. Now whether that’s because the son doesn’t influence family decisions, or whether the son+mother or son+ather are in total alignment so they act like one person, or whether the mother (or father) just does whatever is in his son’s interests, or whether the son is going off and doing his own thing and not telling his parents, I don’t know. Here’s the reference, it’s forthcoming in Economic Journal: http://ideas.repec.org/p/iza/izadps/dp3728.html
One of my personal favourite papers on children and household decision making is this one by Peter Burton, Lori Curtis and Shelley Phipps: http://www.statcan.gc.ca/bsolc/olc-cel/olc-cel?catno=11F0019MIE2005261&lang=eng.
Shelly Lundberg has also done some research on this area. Nothing specifically about minimum wages springs to mind but I have a feeling someone might have looked at something sometime….
Frances: “minimum wages should increase children’s bargaining power”
Minimum wages or higher wages? It’s clearly true for higher wages. What if minimum wages didn’t increase expected income because of higher unemployment/lower productivity? What if a young person only has the skills to perform a job that is worth less than the minimum wage. Then are they trapped?
The minimum wage is a poor anti-poverty program, but I don’t think that is much of the reason for it. Promoting productivity, delineating the dependent from independent, increasing the efficiency of assistance, instilling the value of work pays, are all better reasons to me.
I had missed this bit from this comment by K upthread:
Yet another example of trying to fit a price solution to an income problem.
Very nice. I’m going to have to remember to use that next time.
“The minimum wage is a poor anti-poverty program, but I don’t think that is much of the reason for it. Promoting productivity, delineating the dependent from independent, increasing the efficiency of assistance, instilling the value of work pays, are all better reasons to me.”
I’m not convinced that the minimum wage does any of those things IN PRACTICE. Might a guaranteed minimum income, coupled with serious reforms in welfare provisions, do vastly better in all of those regards (especially diminishing poverty, which is BY FAR the most important)?
Stephen not on topic just some information. Scott Clark & Pete DeVries are now blogging re the joys of anaylizing the some of the information they were responsible for officially reporting and briefing their political master on before they retired from the Dept of Finance. site address: http://3dpolicy.ca/
Marc: when you say “guaranteed income” I assume you are referring to something unconditional (I dont like the term since it sounds like a it’s something means tested or a negative income tax which is why I prefer the unambiguous “citizen’s dividend”)? What problem does a living wage (minimum wage?) solve that isn’t solved better by a guaranteed income?
K “What if minimum wages didn’t increase expected income because of higher unemployment/lower productivity?”
Minimum wages will increase expected income for the more desirable minimum wage workers. For a lot of minimum wage jobs, e.g. retail, sales, I would expect that teenagers/young adults are the most desirable workers, because they’re quicker and stronger and trendier and smarter.
So overall I would expect a higher minimum wage would increase kids’ bargaining power.
Not totally unrelated – remember the post I did about lessons from the animal hospital? One reason that I paid those bills was because I knew that, if I didn’t pay, the kids would.
And where does the money that is distribtued come from? Are these not separate arguments? The assumption is that the state either has the funds and does distribute them? The creation of value is the ultimate foundation of a strong society and that can not come just through subsidy which might be better directed to securing egalatarian conditions for education, health, public transit and the protection of the environemnt?
Thanks for the reply, Frances. Trying to understand your reply got me thinking…
I think I had the wrong mental model of the labour market. In my model, the minimum wage simply set a worker productivity threshold below which workers could not be employed. But if that were the case there would be no particular concentration of wages at the exact minimum wage. I am assuming that the density of workers at the minimum wage is higher than at higher wage levels? How high is it? Is a high concentration of wages at the minimum wage a sign of monopolistic competition: I.e. employers giving up some of their monopoly rents in order to pay above equilibrium wages? Are minimum wages a way to extract monopoly rents? Is there another way to explain the existence (if any) of a high concentration of wages at the minimum wage? Maybe I’m not getting something obvious.
Holy f@#k! What is it with you pricks and your obsessive desire to screw over people who already make a pittance by squashing minimum wage standards.
Heh. Look up, Robert. See the vapor trail? That’s my point, zooming waaaaay over your head..
K, before Ontario increased its minimum wages quite substantially over the last couple of years, about 5 percent of workers were making minimum wage: http://www2.parl.gc.ca/Content/LOP/ResearchPublications/prb0839-e.htm
Then there’s an additional group of workers who make minimum wage +25 cents – a kind of efficiency wage story, whereby employers pay more than the bare minimum to get a slightly higher standard of worker.
About 60 percent of minimum wage workers are female – so as a percentage of the female labour force, min wage workers would be a higher number.
My take has always been that what a minimum wage does, is that it strengthens the bargaining position of the most vulnerable. This holds so long as it is not set to high. But a citizen’s basic income does this better.
The reason for a greater density of workers at the minimum would be a limit in the amount of investment the employer is prepared to make to raise their productivity. An investment or search expense rather than an ongoing subsidy. Anyone requiring a subsidy would soon find themselves unemployed. One has to consider alternatives to the minimum wage. Panhandling can pay more. Prison is much more costly. A guaranteed income is a minimum available to even those who don’t work which would be demotivating to many. There are no perfect solutions.
K,
what you are missing is that at anything but absolutely full employment, the cost of not taking a job for any given potential employee are much greater than the cost of not immediately filling a job for an employer. (Another way of putting it is to say that search costs are crippling for a minimum wage earner, but not for a minimum wage employer.) So usually minimum wage employers will be at a bargaining advantage and will end up paying a wage that is lower than they would have been prepared to pay. Think of yourself and how much you would pay for a babysitter or someone to do the ironing.
Don’t tell me you aren’t guided by standards, but always try to screw out the lowest rate possible (even though you could afford to pay more)! And don’t tell me you always employ the maximum number of low wage employees you could afford at the going rate! I don’t believe you!
It is interesting that on this Canadian (and somewhat more academic) website there seems to be plenty of support for a citizen’s basic income (or like the term better myself – citizen’s dividend). You can’t get anything like that even discussed on American sites. I think part of the reason is the peculiar American political system, which basically doesn’t allow root and branch reform, only single policies which somehow sneak through their change prevention mechanism. Americans basically elect someone to do something, and someone else to stop them. But maybe someone here knows another reason.
Lord:
” A guaranteed income is a minimum available to even those who don’t work which would be demotivating to many.”
I don’t understand this argument. Lets imagine we do it properly with low marginal taxes on low incomes (or just VAT on low incomes). Say the GMI (don’t like the term – try citizen’s dividend) is $6,000 and the lowest wage is $5 an hour. Are you really saying someone will prefer $6,000 and no job to working 30 hours a week (say for 40 weeks a year) and earning 6000 + 30405 = $12000.
And if that is true, would you want to work with this guy? Maybe the society is better off if some people don’t work.
Lord: “A guaranteed income is a minimum available to even those who don’t work which would be demotivating to many.”
You want to motivate people with starvation? Our economy extracts an enormous amount of rent from the natural resources of the earth (in a perfectly competitive economy all profit is land rent). Some people would claim that all citizens have an inherent claim (perhaps even an equal claim) on the intrinsic, natural value of the earth. If so, a citizen’s dividend could pay them the value of that claim. And if some people feel “demotivated” to produce goods in order to increase their own consumption, then that’s just them optimizing their own utility. What do you care? Anyways, it’s a lot less demotivating than welfare with its 100% effective marginal tax rate.
I wouldn’t object myself and would even prefer it, and don’t think everyone need work, but yes, many do want to motivate people with starvation and are opposed to welfare for the same reason. They would prefer charity handle it all, as long as they don’t have to contribute or can dictate where it goes. It is a bit puritanical and controlling with too much emphasis on work but has widespread support. They want both a carrot and a stick and for them the redeeming feature of welfare is the stigma that attaches to it. I don’t see this changing anytime soon.
@ Patrick : Oh, so if those young dudes were in University/College/Trade they wouldn’t be getting drunk? Clearly you’ve never lived near a university.
Despite Patrick, I do agree minimum wage laws don’t really solve poverty issues. And if I’m not mistaken, food-related businesses employ minimum wage quite a bit ( ie grocery stores, fruit farms, etc). If MW was to raise, wouldn’t EVERYONE’S food budget go up?
Mike: You do realize that I wasn’t being entirely serious? I any case, I don’t care what the yobs do with their time and income (such as it is), so long as they do it elsewhere. If lower wages for them means they hold their summer weekend shindigs elsewhere, then three cheers for lower wages!
Please walk me through the personal experience of a poor person who has a minimum wage job for the scenario that
their hourly wage increases from $8 to $12. You seem to be saying, “no effect”, “will still live in poverty”, “waste of money”.
You seem to be implying that there is no point to implement such an increase, not even from the perspective of the
recipient. So please walk me through how such a person’s life would change.
I really don’t know what you would write. So far your discussion sounds cynical. But let’s see if you have some convincing analysis.
Stephan
“Holy f@#k! What is it with you pricks and your obsessive desire to screw over people who already make a pittance by squashing minimum wage standards.”
Clearly, that’s what we’ve all been talking about.
Stephan: For ‘earns minimum wage’, replace any feature that is purely random across the population – say, having a surname with a number of letters that is evenly divisible by 3. Now consider a policy of giving extra money to people with that characteristic. There will be some people in poverty whose welfare is improved, but that would be a matter of random chance than of policy design.
The difference is that no-one (yet) is urgently demanding a policy of giving money to people with last names with a number of letters that is evenly divisible by 3. Nor are they angrily denouncing those who would suggest that such a policy might not be an effective anti-poverty measure.
There seems to be a consensus amongst far-Left groups (International Marxist Tendency, Ginger Project, Communist Party of Canada) to set the minimum wage at about $15/hour. However, most reformist Left groups like the BC NDP, QC solidaire or labour unions seem to want it to be around $10-$11/hour. Most of the reformists aim does seem to be aimed at the income problem rather than price problem.
For example, the Green Party of Ontario’s 2007 platform called for a $10.25 minimum wage – what the government set it to – but wanted welfare benefits set to an urban LICO (raising them from about $7,000 to something like $20,000 a year). The Ontario NDP promised to kick assistance incomes to a committee to come up with livable amounts, while helping lead a $10 minimum wage campaign. Quebec solidaire promised the same thing ($10.25 minimum plus assistance commission) but suggested it might be a guaranteed income near LICO or some poverty measure, I think. QS also wanted to immediately fold social assistance into social solidarity/disability so that incomes rose from about $600 a month to almost $900 in the meantime.
So I think there is a tendency towards raising income supports rather than dramatic future minimum wage increases from the reformist Left. From the centre, the Ontario Liberals contributed to this because most of their anti-poverty strategy has been raising the minimum wage from $6.85 to $10.25 – they’ve basically kept assistance payments at their abysmal Tory levels (except for the child tax benefits) while taking the NDP’s line on minimum wages.
Stephen: “For ‘earns minimum wage’, replace any feature that is purely random across the population.”
Earning minimum wage is not random. Your own data confirms a strong relationship between low wages and low incomes among households with employed members.
However, you calculated percentages by throwing jobless households into the denominator. This approach happens to show that, in Quebec in 2006, the percentage of minimum-wage earners was about the same among low-income households and all households.
That’s interesting, but it does not disprove the relationship between low wages and poverty. It simply underscores joblessness as another important cause of poverty. As I noted above, we need better income-support programs in addition to minimum wages.
Slightly off-topic:
“Yet another example of trying to fit a price solution to an income problem.”
My memory has gone. Who first came up with that line? Was it you Stephen, on hydro prices? Because it’s a damned good way of putting your finger on this and many other cases of bad policy solutions to real problems.
The credit – and it is indeed a credit-worthy way of articulating the problem – goes to Kevin Milligan, who used it to describe the NDP’s proposal for removing the GST/HST from heating. Can’t remember if he said it on twitter or in conversation when he visited Laval a couple of weeks ago.
Erin: A refresher in probability theory seems to be in order. If the conditional probability P(A|B) is equal to the unconditional probability P(A), then A and B are two independent events. In this case, we have
P(Poverty | Earns minimum wage) = 0.135
P(Poverty) = 0.131
From this, we can conclude that earning minimum wage and being in poverty are essentially independent events.
It is true that if you limit attention to an arbitrary subset of those who are in poverty then you get a stronger link. But if we’re going to go that route, then I can claim that a policy of giving money to those whose last names have a number of letters that is evenly divisible by 3 will be extraordinarily effective in reducing poverty among those whose last names have a number of letters that is evenly divisible by 3.
In labour-market policy, employed workers are not “an arbitrary subset.”
Ah. Is this the part where we learn about the importance of making the distinction between the ‘deserving poor’ and the ‘undeserving poor’?
And it’s NOT ‘labour market policy’; it’s redistribution policy. The mistake is using a labour market instrument to do something that it can’t.
We both support higher income transfers to the poor. It does not follow that we should reject every complementary policy other than direct cash payments based on household income.