An overlooked anti-poverty strategy: giving money to poor people

Kevin Milligan of the University of British Columbia and Mark Stabile of the University of Toronto asked themselves "Can income transfers to poor families help children?" Here's a summary of the answer they got:

Since the 1990s, many countries have reformed their systems of transfers to low income families with an eye toward improving work incentives. This column shows, using Canadian data, that well-designed income transfers can not only help families make their way back to employment, but also improve the educational, mental health, and behavioural outcomes of the next generation.

This sort of scholarship makes me despair of the Canadian Left. Instead of focusing attention on targeted transfers, it wastes its time on arguing for higher minimum wages (a policy that is at best pointless), or on making the case for universal programs that favour the well-off. ([1], [2], [3])

When the Canadian Left does consider the idea of targeted income transfers, its response is invariably dismissive. A typical reaction to the Working Income Tax Benefit (our equivalent of the US Earned Income Tax Credit) is to describe it as a handout "so that employers can continue to pay poverty wages." Not all reactions on the left are so negative, of course. But if this piece is anything to go by, the Canadian Left's support for the idea of giving money to poor people can be at best characterised as 'tepid'.

It's getting harder and harder for me to come up with charitable explanations for the disconnect between the policies that the Canadian Left advocates and policies that serious scholarship says will actually help reduce poverty and inequality.

Comments

[Marc]

One of the nice things about minimum wages (and unions for that matter) is they make the labour market do more of the heavy lifting in addressing inequality.

Transfers are great but political support for them is weak ("those lazy ...") -- though recent experience with a carbon tax may provide an excellent vehicle for enhanced income transfers.

Public services are also about what kind of society we want to have, and an inclusive society where everyone has opportunities on an equal basis has virtue to it, not to mention efficiencies (health insurance) and spillovers (education). Otherwise programs designed just for the poor end up being poor programs.

A mix of all of these things seems a good way to go, in my opinion.

[Stephen Gordon]

One of the nice things about minimum wages (and unions for that matter) is they make the labour market do more of the heavy lifting in addressing inequality.

I'm unaware of any evidence to justify such an assertion.

Transfers are great but political support for them is weak

I don't think so - that meme is reserved for people who don't work. The WITB benefits low-income people who work, and it even provides an incentive for low-skilled workers to accept a job. What is disappointing is that the Left won't even make an effort to help such people. And if they don't, who will?

[lickedcat]

I once heard a discussion about a minimum annual income for everyone. Would there be any efficiencies in managing such a system relative to our transfer systems in use today?

I still don't understand the argument for increasing minimum wages. Isn't a minimum wage a signal that the job sucks and that you should be looking for another job?

Doesn't increasing the minimum wage force businesses to increase prices or cut back on hours? If prices increase won't other workers demand a similar percentage or absolute increase in their salaries.

I've always thought the best way to make poor people less poor is to give them money directly. Definitely sending them to social workers or employment offices every week doesn't help them a bit, but sure helps the people offering the "service"

[Stephen Gordon]

I once heard a discussion about a minimum annual income for everyone.

That would be the Basic Income proposal - known as the Guaranteed Annual Income in Canada. Sadly, it too is a very worthy idea that enjoys less-than-enthusiastic support from the conventional Canadian Left.

[Patrick]

I think part of the problem is that you have to actually think a little to see why this works. People like get on their high horse and pump up their egos by pontificating about the moral failings of the poor, and by extension how superior they are. But anyone who has ever experienced poverty can attest to how degrading, how limiting, how spirit crushing it is. Poverty is not a moral failing. No sane person chooses to be poor, and they certainly don't choose to inflict it on their children. It is only the most extraordinary people who can lift themselves out the poverty trap without any help at all. It's simply not right to let everyone who isn't a superstar rot. Providing sufficient income to the poor to lift them out of the morass, so that they can afford to feed themselves, be clean, be properly clothed, so that they can afford a bus ride, and general fit in is NOT creating perverse incentives; it is simply enabling them to be functional in society and thus to be productive.

I can't prove it, but I suspect that the guaranteed annual income would do more to reduce crime than just about anything else you could do (except maybe free drug rehab for anyone who wanted it).

[Caro]

We all pay for poverty, one way or another. We insist on a minimum wage that covers the real cost of living, or we make up the difference by taxing the better off to pay for the less well off, or we just let poor people die in the streets and step over the bodies.

Those are our choices.

Carolyn Kay
MakeThemAccountable.com

[bob]

I agree with this. EITC is definitely one of the better policies we could adopt, and I am likewise baffled that it would be opposed by the Left. Same with the GAI. I was really surprised to see the union objection to that, although I suppose that I should expect such bewildering reactions by now.

The policies on the Canadian Left are pretty reactionary and incoherent. I know that Stephen sees this particular issue differently, but I was surprised when the Lefties who spent 10 years railing against the GST turned around and attacked Harper for lowering it. there just doesn't seem to be any rhyme or reason to their positions.

I strongly favour lump sum transfers as opposed to complex, inefficient social programs. GAI is obviously the first choice, but EITC is pretty much an unalloyed good for working families. I don't see any valid reason to oppose it.

[Matt M]

The problem with targeted cash transfers is that, while increasing the incentive to work in a certain range of the income spectrum, in the phase-out range of the spectrum they decrease incentives to work by increasing the effective marginal income tax rate in this range.

In general, one would think that ceteris parebus decreasing the marginal tax rates for some individuals while increasing them for others (as part of a revenue neutral transfer) would have negative effects on the overall incentives to work. In order to convince me that targeted transfers are a better idea than a per capita disbursement you would need to convince me that the price elasticity of labour supply in the range you wish to transfer to is significantly higher than in the phase-out range.

[Jon]

I am always heartened by list of countries without minimum wages... germany, sweden, norway, finland, iceland, singapore, hong kong... i.e., some of the successful and productive countries in the world.

I recall that in the USA, over 50% of minimum wage earners are minors in middle-class households. On top of this more than 90% of full-time minimum wage workers earn more that minimum wage within one year.

This is actually a poorly considered distortion in studies that attempt to prove minimum wage hikes are beneficial: some such studies fixate on unemployment whereas the wage dynamics are more subtle: when the minimum wage goes up, employers compensate by delaying subsequent wages.

[Jon]

We all pay for poverty, one way or another. We insist on a minimum wage that covers the real cost of living, or we make up the difference by taxing the better off to pay for the less well off, or we just let poor people die in the streets and step over the bodies.

Sounds like a false dichotomy to me. Compound interest is incredibly powerful. Many cases of apparent inequity arise from time-preference differences. One ironic feature is literally that those most desiring present consumption yield the least bounty because of their diminutive stock.

[Don the libertarian Democrat]

"That would be the Basic Income proposal - known as the Guaranteed Annual Income in Canada. Sadly, it too is a very worthy idea that enjoys less-than-enthusiastic support from the conventional Canadian Left."

I am for a Guaranteed Income with health care insurance, as put forward in Charles Murray's "In Our Hands". Is there a Canadian blog that advocates this? I have links to one in the US and one in the UK.

[Adam P]

There seems to be a curious parallel between this post and Nick's post on the Joseph Heath book.

The book, and some of the comments, talked about how the assumption of "self-interest" was such a turn off in basic econ. It seems to me that the rejection of the assumptions of basic econ theory is driving the left's preference for a higher minimum wage.

What I mean is, if you think employers are self interested then in a roughly competitive labour market they should be paying workers their marginal product. In this case the only way to improve the income of those with a low marginal product is to give them what effectively is a transfer, in the sense that you give them more income then the value of their contribution to the productive effort. In this case you want to do it in the way that has the least distrortion to at least maximize the total income that's available for redistribution.

If you don't think labour markets are at least somewhat competitive then you think low wage workers are being exploited, it's not that their contribution is low, it's that they're being robbed by greedy corporations. In this case I think they're arguing for a high minimum wage more to punish the exploiter, deprive him the benefit of the cheap labour, then to help the labourer. This requires them to deny that a higher minimum wage can actually cost poor people their jobs of course but it's consistent with the world view. They think a higher minimum wage is just getting the low wage earner what he anyway deserves.

Afer all, if we imagined a situation where a transfer scheme had exactly the same effect on the actual living standards of the low earners as did the higher minimum wage then someone on the left should be entirely indifferent to which is implemented. Thus, since both theory and evidence imply a non-distortionary scheme of targeted transfers is even better there must be some reason why the left doesn't even want to hear or acknowledge the argument.

[Stephen Gordon]

Interesting point, thanks.

[westslope]

jon, Thanks for sharing the information.

If we were to line up a list of economies with minimum wages and one without, how do income distributions compare?

Just eye-balling the list, I would guess that countries with minimum wage laws would have less equal income distributions (as measured by a gini coefficient, for example) than those without.

[Patrick]

Seems to me that low wage work does not always imply low marginal product, even when market wages are paid. For example, child care workers or cleaning staff are typically low wage work, but their contribution to the productive effort (and public health) or not negligible. Its not as if we can go without child care or cleaning, yet they are consistently paid very low wages. Short of a conspiracy to keep wages low, what does theory have to say about this?

[Matt M]

Patrick, you appear to be confusing marginal with average product.

Marginal product is the benefit from hiring one additional child care worker. Average product is the benefit we currently get from having child care workers divided by the number of child care workers. Due to diminishing returns the second number will be significantly higher than the first.

Because child care workers and cleaning staff require relatively low levels of education there is a large supply of qualified individuals. These services are thus provided in a quantity such that the marginal product is diminished to almost minimum wage level, which is what these workers are paid. If the marginal product was higher than wages then it would behoove an employer to hire more workers and capture the surplus. This would continue until wages had increased and marginal product had decreased until they were equal.

Please note that the above assumes that classical theory actually works perfectly, ignores the tax wedges placed in between producers and consumers, ignores whatever externalities we want to assign to marginal provision of child care, etc

[Jon]

For example, child care workers or cleaning staff are typically low wage work, but their contribution to the productive effort (and public health) or not negligible.

Non-monetary compensation.

[Patrick]

Sorry if I goofed the terminology. I meant marginal product in the sense Adam seems to have meant it when he says "... they should be paying workers their marginal product". I could be misunderstanding Adam, but it I took it to mean the benefit actually realized by hiring them.

I don't agree with the education argument as explaining anything. For example, the average school teacher in NYC has roughly the same level of education as the average hedge fund manager. Precisely zero school teachers earn what a hedge fund manager earns (though school teachers are obvious not poor, it's the disparity that is of interest). Do hedge fund managers REALLY make an economic contribution 2 orders of magnitude larger than school teachers?

[Matt M]

The benefit realized by hiring a single additional child care worker cannot be computed in the manner you suggest, i.e. asking what we would do without ANY child care workers. that is my point. Not your terminology. If there are lots of people to teach school then the marginal product of schoolteachers drops much faster than the average product of schoolteachers

[Matt M]

As for the hedge fund manager /schoolteacher split: if the average schoolteacher could do what a hedge fund manager does then why don't investors hire schoolteachers at much lower wages to manage their investments?

There are some reasonable hypotheses to get to that conclusion, but they're more complicated than trying to say that a schoolteacher with a bachelor's degree "deserves" as much as a hedge fund manager with a bachelor's degree.

[Caro]

>>>>We all pay for poverty, one way or another. We insist on a minimum wage that covers the real cost of living, or we make up the difference by taxing the better off to pay for the less well off, or we just let poor people die in the streets and step over the bodies.

>>Sounds like a false dichotomy to me. Compound interest is incredibly powerful. Many cases of apparent inequity arise from time-preference differences. One ironic feature is literally that those most desiring present consumption yield the least bounty because of their diminutive stock.

Is that English?

My statement was actually a trichotomy, and a description of our choices in dealing with wages. Name another choice, if you can.

Carolyn Kay
MakeThemAccountable.com

[Patrick]

Matt, I think I'm making an entirely different point, but I want to understand what you're saying first. Bear with me as I dredge up some maths I haven't used in a while.

MPL = dQ/dL (pretend the d's are curly and the derivatives are partial)

where Q is f(L, ...).

So if dQ/dL get's smaller as the quantity of labour provided increases, then the slope of Q flattens out as L increases. I suppose that's diminishing returns. And theory says that labour is paid its marginal production (given a bunch of unrealistic assumptions). Average product is easier, so I'll just state is as Q(l)/l where l (little l) is the units of currently available labour and Q(l) is the production achieved at l.

And you're saying that dQ/dL << Q(l)/l. Yes?

Then the only objection I would raise is that I don't think labour willing to do unpleasant work like cleaning (or picking crops etc ...) is really as plentiful as you assume. If it were, then agrobusiness wouldn't seek to import foreign workers to exploit. However, this really isn't my point.

Apparently somewhat incoherently, I was thinking about (low wage) labour relative to other (high wage) labour. Society obviously needs cleaners (for example). I'm going to pick on hedge fund managers, but that's only because they are an easy target. It could just as easily be automotive CEO's.

Anyway ... Without sanitation, life is unpleasant and disease is rampant. It is a necessary and valuable service. It is productive and enables productivity. In contrast, society doesn't really need hedge fund managers (for example) and a case could be made that we may in fact be better off without them (btw I am not saying we don't need finance). Hedge funds are, arguably, not all that productive (at least recently). In any event, I think it's safe to say that hedge fund managers are not 2 or 3 orders of magnitude more productive than cleaners.

Perhaps it is outside the realm of economics and it's purely a value judgment to say that in no state of the world is it justifiable that hedge fund managers be paid 2 or 3 orders of magnitude more than teachers. Personally, I wouldn't want to mess-up the price/wage signals that say "more teachers!" or "no more cleaners", but it seems crazy and wrong that the market signals are REALLY telling us that the world needs millions more hedgies. Something just ain't working.

I was simply wondering if economic theory had anything to say on the subject.

Actually, now that I've written all this, it occurs to me that perhaps there is another interpretation. Supposing Stiglitz (among others) is right (he has a habit of being right) and the world is suffering from a lack of global AD / a savings glut, then maybe in the case of hedge fund managers and CEO salaries the wage signal is really telling us that there is far more capital looking to be productively deployed than can be productively deployed, and thus the demand for people to do it essentially become infinite and so salaries went to the moon ... Ugh.

[Adam P]

Patrick, I haven't read your exchange with Matt all that closeley and I don't want to put words in his mouth but to me the issue is something like this:

When you say sanitation is a necessary and valuable serviec what you're saying is that for sanitation services Q(L) is very large. With this I agree, the issue is that for this service L is also very large and this makes the contribution of any one of the sanitation workrs small. How small? Well, as small as dQ/dL. After all, we usually say dQ/dL is the marginal benefit of hiring one more worker but we could just as easily say it's the marginal benefit we got from the last worker hired.

Or we could say it's the marginal loss of losing one of the workers to another industry and this last formulation is independent of how we order the workers in our indexing scheme. This formulation seems to me to also make clear the distinction, the total contribution to the productive effort of the sanitation industry (equal to Q(L)) may be huge, but the contribution of any one sanitation worker (equal to dQ/dL) may be small.

[Patrick]

I was sloppy in my wording of the last bit of my previous post. I should have said the demand for people to deploy capital shifted way off to the right while the supply didn't change much, so the equilibrium price went way up.

[reason]

Steven,
fully agree, except that I actually think a minimum wage is useful, but also agree that it is important that it is not set too high. And of course if we did have what you call a Guaranteed Minimum Income (my prefered solution) then a minimum wage would be unnecessary (- but you could also call it a "Citizen's Basic Income" or even a "National Dividend"). A minimum wage is necessary if you don't have some sort of basic income support because it equilizes the unfair advantage potential employers have in bargaining with the very poor.

[reason]

" Compound interest is incredibly powerful. Many cases of apparent inequity arise from time-preference differences. "

Yes if I can save $2 a week at 2% (real) I end up after 40 years with 6281.81. Wow, poverty is solved!

(The trouble is, you need money to start with - the rate of return on safe savings is very low, the rich can afford to take more risk).

[reason]

Carlyn Kay,
I guess what Jon is really saying is that the poor would just stop eating, we wouldn't have any problems, they could save the difference.

[reason]

I live in Germany and can say that at present there is a debate about whether to introduce a minimum wage. The reason is complicated (the long border with Poland plays a role), but a lot of the cause is changes to social security and unemployment insurance, put more pressure on the poor to take jobs. I think if you do away with a minimum wage, you need some sort of income support for the poor.

As for incentives to work, I'm not convinced that the evidence says that provided they don't reach a threshold level (c 60%?) or increase dramatically suddenly, marginal rates have relatively small effects on incentive to work. And if you used more consumption and land taxes rather than income taxes (as Steven favours I believe) even this effect is diminished. And don't forget that there are TWO incentive effects of income taxes, an income effect and a substitution effect, and they work in opposite directions. As I see it the income effect is more important here - for those at the bottom of the earnings spectrum. We could expect some people, but the very least productive, to give up working altogether. Perhaps to the relief of their colleagues.

[reason]

The reason that Matt M and Patrick cannot communicate here is that are using different measures of productivity. One is saying that productivity is marginal revenue and the other that it is marginal social utility. Not all social utility is captured by the market.

[reason]

And the reason that hedge managers get paid so much is that they are exploiting the assymetry of incentives created by limited liability and agency problems. If they faced the possibility of having to pay back all the losses they made, nobody would even take the job.

[reason]

Patrick,
or to put it another way, Matt M thinks the needs of rich people are much more important than the needs of poor people and you don't. But both should surely agree that Steven's solution would move your positions closer together.

[Adam P]

reason, I think your 5:09 about externalities is exactly right. But that also explains why it should be a transfer of public money rather than private money from the employer. The employer pays for the marginal benefit that accrues to him, then we all pay for the social benefit that we all share in.

[reason]

Adam,
I don't disagree with you - not just from the point of view of equity, but from the point of view of effectiveness in acchieving our aims. It is always harder if incentives are alligned against our desired goal.

[Matt M]

"or to put it another way, Matt M thinks the needs of rich people are much more important than the needs of poor people and you don't. But both should surely agree that Steven's solution would move your positions closer together."

For somebody named "reason" you're certainly engaging in some leaps of logic here.

I have made no reference to need in any of my comments. Nor have I made any reference, other than deride the concept, to desert.

I am wholly in favour of redistributing income. If I wasn't much more concerned with the well-being of the poor than I am of the rich then I would be busy decrying the entire concept rather than trying to argue over the best way to go about doing it.

As for the supposed large positive externalities you're trying to claim derive from the labour of marginal low-paid workers, could you please do us the favour of at least attempting to enumerate some of the most important ones? Do you think that these large positive externalities apply across the board, or do you have some favourite occupations which you think should be subsidized?

[Caro]

>>I guess what Jon is really saying is that the poor would just stop eating

Choice #3 - let them die in the streets, and step over the bodies.

Carolyn Kay
MakeThemAccountable.com

[reason]

Actually Matt,
no I was just pointing out the consequences of your logic, in stark terms (which a lot of people don't like clearly). The market provides what people with money want, it ignores the desires of people without purchasing power. Patrick was measuring "usefulness" with a measure that treated all utility as valuable, whether it was monetised or not. And surely, you must agree that Steven's solution brings you closer to one another - I never actually said you wouldn't agree with Steven's solution. And as for quantifying externalities - it suffers from the same weighting problem. What units should I use?

[reason]

And it always seems to me, while we are on this topic, there is something wrong with this marginal/average wage topic. The problem is that it is not possible to calculate the marginal productivity of any given worker. Every one is employed sequentially and usually work as a team. What would there contribution be if they were the last one employed? Would it be different from another one - and if so why should their pay be governed by that of the least productive member of the team? (And I won't go into increasing returns to scale and imperfect competition). The logic of the argument seems to imply to me, that significant involuntary employment, implies an open ended fall in wages. Given the inflationary nature of full employment (over the natural rate) it seems macro-economic managers must walk an impossibly narrow tightrope.

[Adam P]

reason: "it is not possible to calculate the marginal productivity of any given worker. Every one is employed sequentially and usually work as a team. What would there contribution be if they were the last one employed?"
I sort of tried to take care of this in the last paragraph of my 1:54 above. But nothing there rules out "synergies" or assumes all team members have the same marginal product. It's quite common to assume increasing marginal product when few are employed at a task exactly because the team effort makes everyone more productive. That doesn't mean that when many are employed at the task their marginal product doesn't decline.

[reason]

Adam,
I understand that, and I understand the argument that produces wages = marginal revenue earned for the production of the last worker employed (i.e. that if is less, why isn't another one employed). The problem is what does this marginal logic, really mean in agregate and in markets that aren't in equilibrium, and where workers are not homogenous and there are information asymmetries. If there is an excess supply of labour in any market, then employers can keep reducing the wage offered, until the excess supply finds better paying jobs elsewhere. The marginal worker, is in reality in competition with other marginal workers, and the marginal revenue he can earn is not known. So it is at best a first approximation, to a much more complicated story. But it doesn't change the fact that it is marginal revenue that is the key to what employers can pay, so the actual social utility of what is done is irrelevant, it is what people are prepared to pay for the product that counts. And it is not just externalities that matter here, it is also consumer and/or producer surplus.

[reason]

Oh, and just another little aside - shouldn't people stop talking about marginal productivity and start talking instead about marginal revenue. It would save a lot of confusion. Productivity really is a lousily ill-defined concept - particularly in a service oriented economy.