One rationale for the public provision of public goods

Amidst all the current talk about nationalization, it is worthwhile to think when the provision of private goods by public entities is indicated. Generally the argument is that the good is a natural monopoly and the particular industry is difficult to regulate. But are there other reasons?

Hanming Fang and Peter Norman provide an intriguing alternative explanation: when a government provides a private good, it learns more about the consumer and is thus better able to tax in the sense of optimal taxation. The basic idea is that it is difficult to configure properly a tax system because of information problems. In this particular case, there is private information about how much households value different goods, and the government tries to determine the optimal price à la Ramsey pricing. In addition, Fang and Norman show that the optimal pricing would call for a tax on the private good to subsidize the public good. This tax would be dependent on whether the household also consumes the public good.

While this is an interesting argument, one can question how this translate into practice. The authors suggest that the government should bundle the goods. This is something that is done aplenty in the private sector (think of bundling phone, cable TV and internet access), but this is done for similar goods. I cannot think of examples where public and private goods could be bundled in a reasonable way and that would be informative.

How to increase employment, and at what cost

Governments across the world have been scratching their head on how to increase employment, or decrease unemployment, and the question appears to become more urgent these days. History shows that they have been extremely creative, but with little impact. The main problem is that they have tried to force some people to work less in order to spread the total number of work hours among more people, as if this were a zero sum game. Particular fiascos are the French 35 hour week and countless early retirement programs.

Economists generally do not think it is appropriate to add constraints to a market that is not in equilibrium. One should rather work through prices by imposing taxes or subsidies in order to encourage people to take some actions while losing little efficiency. The question is then what to tax and what to subsidize. Victoria Osuna takes here an interesting perspective with a model economy incorporating commuting, team work and overtime. There are three policy tools: tax overtime, subsidize wages and subsidize employment. All make work cheaper, but the substitution of hours in the workweek becomes critical. This is embedded in a real business cycle model, thus general equilibrium effects influence capital accumulation and factor prices.

The experiments are quite interesting. First, tax overtime to bring hours from 40 to 35: you need a 12% tax, steady-state employment increases 7%, but GDP decreases 10.2%. The issue is that a lot of capital now lies idle. To get the same increase in employemnt, a wage subsidy is desastrous, as it reduces GDP by 12.7%. The problem is that the subsidy needs to be substantial, 16.5%, which leads to large misallocations of capital. An employment subsidy of 4.5% does better, dropping GDP by "only" 4.7%.

So, if you want to increase employment, be aware that there will be substantial costs in terms of output and lost productivity. I am not sure it is worth it.

Self-confidence and self-employment

Who decides to become self-employed? This can be an important question if one considers that entrepreurship is a major factor in economic growth, and entrepreneurs are largely self-employed. Ozkan Eren addresses this question using the US National Education Longitudinal Study data and comes to some surprising conclusions.

The extant literature has firmly established that wealth and the father's labor market choice are very important determinants in self-employment choices. And this seems to make sense: you need to be capable of overcoming financial hurdles and have a role model to follow. This literature neglected, however, the impact of cognitive and especially non-cognitive abilities. NELS data include measures at 8th and 12th grades like test scores, self-esteem and locus of control that Eren exploits here.

While it may not seem surprising that individuals with higher self-esteem tend to be more self-employed, but it is counter-intuitive that the impact of cognitive skills is negative. One can rationalize this, though, with the idea that high skills can be compensated with wages, and lower skilled people may find fewer opportunities on the labor market and work by themselves. Interestingly, the two abilities work rather independently, as only one fifth of the correlation between cognitive abilities and self-employemnt can be explained by non-cognitive abilities.

Americans stand out against Europeans because they exude self-confidence. This may explain why there are more entrepreneurs in the US and why income a higher, despite the fact that US test scores on abilities are lower.

Democracy and public good provision

Is democracy good for an economy? As discussed previously on this blog, there does not seem to be a link between income and democracy. But this does not mean that the political regime does not have an impact on an economy. For example, a democracy may care more about the provision of public goods.

This is what Sebastian Vollmer and Maria Ziegler explore empirically. Using a panel data analysis, they find that life expectancy and literacy are positiviely affected by democracy. This can be explained that as schools and health are publicly provided in most countries, the major dimension of change is how much (not if) governments care about schools and health, and democracies care more. Of course, there are always exceptions, such as the US where the government does not care that much and it is reflected in outcomes (after controlling for GDP, of course).

If democracies care so much, why are they not richer? It may be because redistribution, which the provision of public goods essentially is, is not necessarily growth enhancing. Autocracies care only for the elite (but not too much, or the common people will revolt), and the elite's income is mostly driven by aggregate GDP. But the large level of rent seeking in such economies drives GDP down. The point of my rambling here is that GDP should not be the focus of the analysis anyway. Thus, inequality and redistribution will matter for social welfare, and democracies appear to deliver this.

Minimum wages and optimal taxes

The analysis of the impact of minimum wages is usually limited to empirical studies of the employment or unemployment elasticity. Theory is believed to be rather straightforward. In this respect, the work of David Lee and Emannuel Saez is refreshing. They study the interaction of a binding minimum wage with income tax policy and find that minimum wages are beneficial as long as taxes are sufficiently redistributive and the resulting rationing is efficient: the workers with the lowest surplus do not get jobs.

The analysis is perfomed within the context of optimal taxation theory and allows for non-linear taxation. Then, minimum wages essentilally become part of tax policy and they allow to disciminate by labor efficiency. This is essentially the argument that Pierre Cahuc and Guy Laroque have made: minimum wages become redundant once optimal taxes are in places, or you can assume minimum wages and optimal taxes will work around them.

The two papers have an important distinction. The first assumes perfect competition, the second monopolistic competition. As argued before on this blog, minimum wages can lead to monopolistic competition due to implicit collusion that they favor. So all in all, I am not sure the Lee and Saez results are that exciting...

Gender based taxation

If governments need to raise revenue, it is well known that the best thing they can do is through sin taxes, such as for carbon emissions in general, gas in particular, alcohol, tobacco or workalcoholism (or subsidize fitness). If this is not sufficient, the most inelastic goods should be taxed, i. e., goods whose quantity is little influenced by changes in its after tax price. In the context of labor income taxation, this means men should be taxed more than women. Indeed, the female labor supply is much more responsive to after tax wages, as men work anyway.

Alberto Alesina, Andrea Ichino and Loukas Karabarbounis expand on this idea, noting first that is common to discriminate by gender, say through affirmative action, different retirement policies or maternal leaves. So why not discriminate more efficiently through the price system instead of quotas? From an economic point of view, this seems obvious. But somehow this is not popular with the general public. Is it because the basic economic model is missing something?

Alesina, Ichino and Karabarbounis test the robustness of gender based taxation to all sort of bells and whistles, and its optimality remains. In particular, they assume that men and women have identical characteristics and preferences but for the fact that men have higher bargaining power at home and can thus avoid household chores by working in the market. As long as household chores are unbalanced, taxing men makes more sense and improves overall welfare.

That said, there is no reason to stop at gender. I have argued before that luck at birth should be taxed, such as for tall people. Others have made similar argument for beauty. In fact, the strongest argument for development aid can be made on the grounds that where you are born is not something you have earned, and one should share one's luck with unlucky ones. And your birth place certainly is inelastic.

Subprime borrowers make stupid financial decisions

Much has been written and said about some ill-informed decisions that subprime borrowers have taken with respect to the housing market, in particular taking mortgages that they cannot afford in the longer run. Some of this has been attributed to being ill-informed about the terms of the loans rather than irrationality. Mortgages can be complex contracts with lots of fine print. What about more straightforward financial instruments?

Sumit Agarwal, Paige Skiba and Jeremy Tobacman merge data from a credit card company and a payday loaner, the latter often charging interest rates that annualize to several hundred percents. They notice that people borrow from payday loaners while still having significant liquidity of their credit cards. An amazing two-thirds of those taking a pay-day loan have more than $1000 available on their credit card, and the loan they are taking averages at $300. Over the two week life of such a loan, $52 would have been saved by putting it on the credit card. Considering that payday loans are often repeat businees, we are talking about substancial amounts.

I reported earlier on the puzzle that some people have credit card debt while also having significant amounts in checking accounts. That puzzle could be explained by liquidity needs for transactions that need to be paid by cash. But this new puzzle is really baffling: both credit card and payday loan provide the same service at a very different price, and people use both even if unconstrained. Are people really that stupid? And so many people?