Confusion on income and poverty

I have heard this sort of claim quite a bit from friends in recent months:

Doesn’t that sound grand – if the richest 100 people in the world gave up a quarter of their income then SLAM poverty gone.  Ez.

However, this isn’t quite right.  In fact it is very much not right.

Instead of thinking about quantities of paper money, we need to think about the underlying claim on real goods and services this presupposes – what exactly changes when we shift income this way.

Cut the income of the wealthiest and they will invest less and buy fewer nifty leisure goods (say boats).  Give this income to the poorest in the world (if that is possible, as in many of those cases people are stuck under a corrupt regime trying to extract anything) and they will demand food, water, and clothing.

And here is the little secret economists keep trying to make public, prices will change.  Without reference to a “set” of prices, income comparisons make no real sense. (Note:  For those who will note it, changes to labour supply would also occur in all of this, the wage is a “price”).

When I did the series on tax, especially here, I was trying to point out that if we try to do “large” changes to the structure of an economy, we are going to have much larger sets of unintended changes – many of the simulation and structural models economists use, even with all their complication, can often only sensibly discuss “small” changes in tax and benefit structure.  Trying to use a naive estimate (note this isn’t an insult, it is a technical term for modeling in the way suggested by the twitter quote) struggles even more with these changes in prices and the real values of incomes.

Many people who believe that we can take money off the rich to feed the poor simultaneously believe we are “stretching the limits” of the earth’s ability to produce food and other goods and services (such that the ‘costs’ of making more of these necessities will rise strongly if we aim to boost production significantly).  These two views implicitly contradict, and we can only understand and try to measure the true trade-off with an honest view of the change in prices!

Now don’t get me wrong, income transfers should still change the underlying production and claim on resources!  But how do they help?  By increasing the relative return (in terms of goods and services) for the thing the poor desire, and by decreasing the relative return for those who create boats and handbags and the such.

It is also a balancing act though – by trying to lower the income of those who do invest, we also reduce capital investment overall.  We change its composition and its level.  Furthermore, when we provide an endowment of income we change the incentive to supply labour.  This requires a view on societies underlying ability to produce, the underlying “production frontier” that exists, the way the “production frontier” moves, and the way relative prices and endowments will adjust as a result of these policies.  This is much more complicated – and also much more fundamental to actually understanding issues of inequality, opportunity, and capability – than what Max’s arithmetic example stated in the above tweet.

And I believe we have heard what this is before – it is the equity-efficiency trade-off from redistributing goods and services.  What do you know!

Note:  This is not a criticism of Max – it was just his post I spotted that helped bring up the issue.  I see this mistake happen CONSTANTLY, it can even be done inadvertently by economists I take incredibly seriously and find very insightful:

But it is still a relatively fundamental error.

Sidenote:  The World Bank discusses why their focus is on alleviating poverty rather than focusing on inequality when it comes to development economics here.

Update:  Dr Phil of Economics notes that, when it comes to global poverty, it is the messiness of institutions that is the most relevant issue – beyond what is described here:

  • jamesz

    If I read you right you’re saying that massive redistribution ‘changes other stuff’. That seems pretty uncontroversial. The bigger question is how much that stuff changes. Is it enough to offset any gain from the redistribution? Half of the gain? Saying that we don’t know because only small changes can be estimated isn’t a very strong argument against attempting redistribution.

    • Matt Nolan

      I’m saying that prices are endogenous, and that just shifting money incomes doesn’t tell us how claims on resources change – and that this problem gets significantly larger the larger the shift in income is.

      This is incredibly uncontroversial … for an economist. For many non-economists this is wildly unintuitive, and it is directly what is stated in the post :)

      “Saying that we don’t know because only small changes can be estimated
      isn’t a very strong argument against attempting redistribution.”

      We know the difference of the efficacy of redistribution though – it is smaller and its impact on solving real issues of poverty are then both smaller and more uncertain. The claim in the tweets is that a certain transfer will “solve poverty” – when such a claim actually lacks teeth.

      On a global scale, it may be better to focus directly on providing access to goods and services that the poor require, rather than arbitrary measures of money incomes.

      Note that, on a national level, this price issue is significantly less binding – due to the ability to trade. In that case, the issue becomes one that is less about those prices, and an equity-efficiency trade-off is easier (still hard) to work out.

      • Jason B.

        Fair post, but to challenge you on this: “On a global scale, it may be better to focus directly on providing access to goods and services that the poor require, rather than
        arbitrary measures of money incomes.”

        I’m pretty sure conventional wisdom is that providing goods and services is the best form of aid. And this has been challenged recently by experiments involving unconditional cash transfers (see: ). From what I’ve read, the idea of directly transferring income (well to be fair, the source of the money was not specified in the experiments. It’s just purchasing power. Could come from donations, or tax, or …) is the new kid on the block in development economics.

        • Matt Nolan

          Indeedy – one of the key reasons I wrote “may” was with that directly in mind. Furthermore, I think I would prefer a situation where we transfer money incomes rather than trying to “pick” goods and services – so in that regard I think I misspoke in comments ;)

          But let me discuss it in any case. One thing to keep in mind is that even with those schemes, the impact is expected to change as programmes are scaled up. Again if we are simply looking at a shift in money incomes from the very wealthiest group to the very poorest over a global economy there will be a shift in prices that does not occur with small targeted transfers. As a result, this does not suggest that the impact on poverty will just “scale up” either.

          This is most certainly not to suggest that it is the wrong way – if we could solve institutional problems, then money transfers likely offer the best way forward, as it does not presuppose the goods and services that the poorest desire to push them out of “poverty” as they would define it subjectively.

          The key reason I think I wrote that line is because, in my mind, I was
          thinking about “the ability to buy a certain minimum level of goods and
          services” as the key thing, rather than having money income at
          pre-transfer prices – so I was thinking of it in terms of providing the
          capability to reach a certain minimum level of consumption. When I read
          it now it seems like I am saying “intervene in specific markets” –
          which is indeed a touch strange for me to say!

          My key point in all of this is simply to point out that prices would change, and as a result those types of comments that are fairly common in public (if so and so gave up their income we’d feed XXX people) are a touch off base – even if they are well intentioned!

  • boristhefrog

    I guess there is also the ‘small’ definitional issues of what poverty is trying to be cured… a poor person in NZ is very much richer than a poor person in Kenya, for example…

    • Matt Nolan

      The argument is about absolute poverty in a global sense indeed!

      The argument about either absolute or relative poverty within a nation is a touch different, especially a small open economy, since the impact on prices will not be as significant.