The idea of a basic income - it's generally referred to as a Guaranteed Annual Income in Canada - has been floated again. My initial reaction was that this is such a good idea that it's hard to figure out why we don't have it already.
For some reason, there doesn't seem to much in the way of a formal modeling exercises that work through the general equilibrium implications of a BI (I will be happy to be corrected on this point), so I decided to set up a blog-sized version. It turns out that although the BI is still a good idea, it's not quite the slam-dunk I thought it was.
The gory details are below the fold.
The model is the simplest one I could imagine: labour is the only factor of production, and since there's no capital, the model is also static. As in any modeling exercise, I'm not pretending that these assumptions are realistic; my goal is to get some basic understanding of the interactions between redistribution policies, labour supply and welfare. The model is calibrated to reproduce a median income of $25,000/year with a gini coefficient of 0.52, which is not far from what we have now. The details of the model and how it was calibrated are written up here.
Two policy experiments were considered:
- Conventional welfare, in which agents have access to a minimum income, but those who choose to accept it are not permitted to work. Those who work are subject to a flat income tax.
- A basic income, in which payments are not conditional on the labour supply decision, and in which only market income is subject to a flat income tax.
In the graphs that follow, the horizontal axis is the rate at which market income is taxed. The range of variation of the tax rates corresponds to values in which it is possible to satisfy the zero-deficit condition.
First up is labour participation rates:
This illustrates one of the strongest arguments in favour of a BI, and against conventional welfare: conventional welfare forces marginal workers out of the labour market. For any tax rate, an economy with a BI will have higher participation rates. Note also that participation rates fall below 0.5 pretty quickly.
Next up is average income per capita:
Because conventional welfare forces workers out of the labour force, total income - and hence average income - is generally lower than the BI case. (I'm not sure what's happening at the very left-hand side of that graph).
Next up is income inequality:
Since a BI benefits everyone in the lower part of the income distribution - and not just those who choose to exit the labour force - a BI reduces income inequality more than does a conventional welfare program. There is a region in which conventional welfare reduces inequality more than does a BI, but as we will see shortly, we don't want to go there.
So far, the advantages of a BI look pretty clear. But the analysis is not so clean as all that. Here is a graph of payments against tax rates:
Since the payments are targeted under conventional welfare, they can be set at higher levels, and the gap is highest for low levels of taxation.
This matters when we look at welfare effects. The metric I'm using here is average utility across the population:
It is here that we see that the advantages of the basic income are not as clear-cut as all that. For low levels of taxation, conventional welfare provides better value: public funds are concentrated at the left-hand tail of the income distribution, and since since those workers are the least productive, total and average income isn't much affected by their absence from the labour market.
It is important to note than in both cases, there is a point at which increasing payments reduces aggregate welfare: the costs of seeing more people leave the labour market exceed the welfare gains of increasing the incomes of those at the left-hand tail of the distribution. Here is a graph of welfare against payments expressed as a percentage of average income:
(This graph takes into account the negative effect of higher payments on average income note earlier.) Curiously enough, the optimal payment in both cases works out to about 50% of average income. This may be a coincidence, or something built into the model; I don't know. But it should serve as a warning to some of the more excitable supporters of a basic income: payments will not necessarily be all that generous.
It is the case the the BI optimum is superior to the optimum we can obtain with conventional welfare, but attaining this optimum is contingent on taxation levels that are higher than what we have now.
This is hardly the last word, of course. A more realistic model would incorporate physical and human capital:
- If human capital is endogenous, then the effect of income taxes on output will be accentuated. The reduced rate of return on investments in human capital will lower output and income.
- A model with physical capital and savings introduces an alternative mechanism for generating tax revenues: consumption taxes. Substituting consumption taxes for income taxes will attenuate the disincentive effects of a BI.
The case for a basic income is still a strong one, but someone will have to work through these issues before it becomes a serious proposal. And BI proponents should also be prepared to make the case for higher levels of taxation.