Online purchases and state use taxes

Most countries by now have some sort of consumption tax. In the United States, this is a state-level tax collected at the retailing point, a sales tax applied in most states. This intention is to apply the sales tax to the state's residents, not its retailers. Purchases made in other states need thus to be taxed in some way, and this is typically done in the annual income tax return by filing a use tax form. Very few people keep track of their cross-state purchases, and even fewer comply with the filing. Does the same apply to online purchases?

James Alm and Mikhail Melnik use data from eBay and find that 94% of all purchases are cross-state. What they should do is compare this with the probability that a purchaser who does not take into account the state of sale. With such a random matching, my rapid calculations give 95.7% (I included the District of Columbia). Thus 94% indicates that eBay buyers do have a preference for the home state, unlike what the authors state. A missed opportunity for the authors who then make a strange statement: "The average state online contribution is 1.66 percent of all purchases on eBay.com." I could have sworn that there are 50 states, and thus each one's average contribution would be 2%.

The paper does not get better after that. We learn that the correlation between state population and sale volume is close to one. One somewhat interesting number is the share of sales tax potentially attributable to eBay, 0.43%. That number is an extrapolation from the consumer electronics sales on a single day in 2007, thus quite uncertain (also because the same sales tax rates was used for all states). The author claim this number is small. When you consider that eBay represents one eighth of online sales, revenue from use tax should be about 3.3% of sales tax revenue. Not negligible. But Alm and Melnik do not bother looking up actual use tax revenue. Too bad.

People do not want a cashless society

With the technology available today, we would be ready to switch to a cashless economy. While having every transaction recorded in some way may trigger some privacy issues, imagine how easy it would be to keep records, file taxes, avoid theft, and how difficult tax evasion would become. So are we ready?

No, argue Naoki Wakamori and Angelika Welte, and it is not because merchants are holding back. Household still hold on dearly to cash, at least for small transactions. Indeed, the Bank of Canada conducted three years ago a survey on payment methods. One can use its data to estimate a model of payment method choices, then run a counterfactual experiment where card payments would be available universally. Wakamori and Welte find that cash usage barely changes. This means that households have a strong preference for using cash. Younger generations, though appear to be more ready to use debit or credit cards, so a cashless society may still lie ahead.

The impact of irrigation on democratization

There is a large variety of institutions and cultures around the world, and I find it fascinating to understand how they originated. The literature has shown that some of them have origins that go back very far, possibly all the way to the introduction of agriculture (Example 1 and example 2)

Jeanet Sinding Bentzen, Nicolai Kaarsen and Asger Moll Wingender discuss how irrigation has influenced the arrival of democracy, and in fact prevented it. They point out that this is not a story about irrigated versus rain-fed agriculture, but rather of areas where there was potential for irrigation versus the others. The story is as follows. Irrigation requires large investments and control of water sources, and allows the extraction of substantial rents. This is a perfect environment for a despot. This theory was advanced by Wittfogel in Oriental Despotism and was soundly criticized. But Sinding Bentzen, Kaarsen and Moll Wingender confirm that the theory holds water empirically. The fact that irrigation potential is crucial allows to avoid the reverse causality issue, as geographic factors are presumably exogenous to institutions, at least until man started having an impact on climate. What is fascinating is that this still has an impact today on the prevalence of democracy.

What makes the wife happier?

The flexible margin in a household's labor supply is usually the wife. As long as the husband has a decent income, the decision to work for the wife can then hinge on more than strictly economic factors, such as finding it more satisfying to work or not to work. Everyone is different in this regard, yet there may be some aggregate trends that may help us better understand the female labor supply as well as what can improve the population's well-being.

Edsel Beja finds that women are generally just as happy being a housewife or a working wife. But there are subtleties once you look closer at the data from the World Values Survey. In middle-income countries, there is a clear ranking in satisfaction: part-time work is best, housewife second, and full-time work is last. This should contribute to the realization that work is not an all-or-nothing decision, and that flexible work schedules can be very much appreciated.

Why you want good national statistics

In this age of austerity measures, every public expense is subject to higher scrutiny than usual. Statistical offices are under particular pressure as politicians do not quite seem to see the point of making informed policy decisions (or are actively trying to hide the real situation of the economy, as in Argentina and possibly China again). What makes good public statistics valuable? One could make a theoretical argument about optimal policies, policy uncertainty and expectation formation, but here nothing beats empirical evidence.

Oasis Kodila-Tedika looks at Sub-Saharian African countries. Of course, these results may not carry over to developed economies, but Africa has the advantage of bringing substantial variation in the data to the table. He uses a measure of statistical capacity from the World Bank and shows that higher values have a positive impact on government effectiveness, as measured by survey results on perceptions of governance. While the results are statistically very significant, it is impossible, though, to determine whether they are economically significant, as the units in the variables have no direct economic meaning. But an improvement of statistical capacity from the worst country to the best one would improve more government effectiveness than improving GDP per capita from the worst to the best. That is huge, even though there is substantial endogeneity that needs to be accounted for, but I doubt it would make all of the effect disappear.

How to encourage research into neglected diseases

There a few things that infuriate me about the pharmaceutical industry: the obscene returns on investment it gets, how it games the patent system, and how it neglects some important infectious diseases. But the pharmaceutical companies just respond to a poor institutional setup, where patents are too generously given and governments favor them too much compared to consumers (and the government itself, as in the USA). As for neglected diseases there is simply no money to be made as these are limited to poor developing countries.

Frank Müller-Langer studies how the discovery and provision of drug preventing and treating currently neglected infectious diseases could be improved. He distinguishes push and pull subsidies. Push programs are those that subsidizes the inputs, such as lab equipment or manpower. Pull programs subsidize results. The question is what to use. Basic research should quite obviously be helped with push programs. This is typically done in an academic environment where rewards come from publication, all you need is the means to conduct research, the will is there. The next phase, development and implementation, should more be helped with pull programs. Incentivizing results is more important there and could lead to surprises. I am thinking here, in a very different context, about the Ansari X Prize that rewarded with $10 million the first non-governmental organization to successfully launch a space shuttle. The prize seemed low compared to the costs, yet 26 organizations participated. Something similar should work for malaria, yellow fever, etc.

The impact of the minimum wage in a developing economy

Whether raising the minimum wage decreases employment has always been controversial, foremost because some studies could not find such an effect, second because there are some theoretical grounds that can support the opposite effect. Some of this has been discussed on this blog: 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11. Only one of those works pertained to a developing country, where it is found, using household level data, that an increase in the minimum wage leads to a shift from the formal to the informal sector and leads to a decrease in poverty.

Ximena Del Carpio, Ha Nguyen and Liang Choon Wang provide new evidence, this time from Indonesia and using data from manufacturing firms. The impact of an increase of the minimum wage depends highly on what you are looking at: it increases employment at the provincial level, but decreases it at the firm level. The negative effect is particularly strong for non-production workers, workers with few skills, female workers, and small firms. It should not be a surprise that people with lower skills are more affected, and small firms, too, as they tend to have lower wages. But this just confirms that empirical works on the minimum wage are hard and confusing.

There are too many swimming and gymnastics medals

As the 2012 Summer Olympics draw to a close, people get all upset about how the medal count does not truly reflect which country is the best because some events are not true sports, some other is over-represented or absent, or judged events do not count and team events should count more. This is all bullshit. First, the International Olympic Committee does not even have an official medal count. Second, any such discussion makes no sense if one does not at least take into account the size of the country. Nations with hundreds of millions of citizens will obviously have a better shot at finding star athletes in many disciplines, let alone find several of them for a team event. To take an example, that the Bahamas can be a medal contender in track relays is in my mind a much more valuable outcome than the US winning a handful of track gold medals. If anything, countries should be ranked by how much better they contended compared to what one would have expected from their size and development. The larger the positive forecast error, the better.

So much for the first part of my rant. The second is that some events are crassly over-represented, and here I am pointing my finger at gymnastics and swimming. My criterion here is the ease with which a single athlete can amass a multitude of medals. I can understand that when there is an exceptional athlete, it is OK for her to accumulate many distinctions. But when this happens repeatedly for a sport, there is a problem. Looking at the list of multiple medal winners on Wikipedia, it is striking how frequently gymnastics and swimming appear. For athletes with ten or more medals, we have 13 in gymnastics, 9 in swimming, 3 in cross-country skiing, 2 in track and field, 2 in fencing, 1 in shooting, 1 in canoeing, and 1 in biathlon, for a total of 32. No matter what statistical model you use, it is impossible to argue that the concentration of athletic outliers can be that high in gymnastics and swimming.

Flexible-price inflation and monetary policy

Are prices flexible or not? There is no doubt that there is some rigidity. But does it actually matter? This debate in the macroeconomic literature has surprisingly side-stepped an important aspect of price rigidity: not all prices are equally rigid. While this has been empirically demonstrated many times, it never really made it into a model.

Stephen Millard and Tom O’Grady take a standard two-sector model and label one sector "sticky-price" and the other "flexible-price." They go through the usual motions of assessing their model and find that it can account for much of what is going on in the economy. But more interesting is their idea that there is a lot of useful information in looking at the inflation rates in the two sectors. Flexible prices react faster to current output gaps, while sticky ones should contain more information about expectations on future conditions, in particular inflation. And it should not be too hard to compute these statistics with current data collection. We already have several works that categorize goods by price stickiness.

Unions just cannot help it

The role of labor unions is to counteract the market power of oligopolistic employers. That is certainly beneficial, but by their very nature, unions have also a perverse effect. As they only represent their current members, they tend to keep employment lower than what would be optimal. Is this the only perverse effect?

Per Krusell and Leena Rudanko take a search and matching model and give unions their best shot: they also care about the unemployed and take into account the impact of their wage demands on job creation. If the union is able to commit on all future wages, we get the efficient outcome. However, if the union cannot commit, then unemployment will be higher than efficient. The reason is that the unions wants to extract rents from current matches. This optimal, time-inconsistent policy can have a substantial impact, three percentage points in unemployment (and GDP). Interestingly, the model also shows that real wage stickiness results form this lack of commitment, as well as excessive labor market volatility.

Are religion and national identity substitutes?

"Religion is the opium of the people," as Karl Marx said. If religion becomes less important, does national sentiment take over? It was certainly the hope of communist regimes that allegiance to party and country would benefit from repressing religion.

Kenneth Harttgen and Matthias Opfingeruse the World Values Survey to construct an indicator of national identity based on questions about social and political participation and political interest, then applying principal component analysis and calling the resulting principal component "national identity." I have never been particularly fond of this technique, as the naming of the factors is very subjective. For example, in this case it could just as well be called "sense of civic duty," "social participation," or "confidence in politics," none of which necessarily tie in with national pride. In fact, I would say that in many cases where voter turnout is high or people sign petitions is where people are rather upset about where they live. Also, the few questions about country and nationality have a very low weight in the index (total of 18%).

Thus, I am not convinced this indicator measures national identity. But let us assume it does. Harttgen and Opfinger then try and see whether it can be explained by measures of ethnic and religious diversity or polarization. The only one that sticks is religious diversity. Does this mean that religiosity is then a substitute for national identity? The authors think so. It is useful to think here what this "religious diversity" variable measures: the probability that two randomly drawn people belong to the same religious group. I do not see how this would relate to religiosity, that is, the strength of one's attachment to religion. Therefore, one cannot conclude in any way from this study whether religion and national identity are substitutes.

The power of check-up reminders

I think dental checkups are the best deal in terms of dental health. They cost relatively little, allow to detect dental problems early and before they turn into root canal procedures, and the dental cleaning gives me the impression that I have new teeth. Yet, because they are rather infrequent (every six months), one tends to forget about them or postpone them. The, how effective are reminders?

Steffen Altmann and Christian Traxler did a little field experiment, sending to clients of a German dentist no reminder, a neutral reminder or a reminder with an explanation of the benefits of a dental checkup. It turns out the reminder is very effective, and a neutral one is just as good as a "promotional" one. Indeed, people know checkups are a good deal, at least people who respond to reminders. No particular difference is detected between those with and without dental insurance.Interestingly, the impact of the reminder is so strong that it is still significant 100 days after sending it: Sending one is enough, people must be keeping it next to the telephone.

How do you want your pain?

Suppose you have some illness and can choose two treatments. With the first, pain gradually increases. With the second, pain is at its peak initially then decreases. Which one would you choose? If your are discounting future utility, I guess you would want to have pain later. But knowing that things will only get worse can be quite depressing. So it is not obvious what I would choose, and I think that choice would be difficult for many people.

Eike Kroll, Judith Trarbach and Bodo Vogt actually executed an experiment with such a choice, including real consequences. Specifically, people would have to put their hands in water bowls at successive temperatures of 12C, 8C, and 4C (or the reverse order) for one minute in each. Before that, they are asked to rate on a scale their preferences among sequences along with a willingness to pay to avoid the worse one. A clear majority of subjects chose the improving sequence, with a two-point spread for the median displeasure points, which is considered significant in behavioral research. As for the willingness to pay, the median is zero (and possible increments where tiny 20 cents). Thus: the displeasure point scale is rubbish, and people may consider more than discounting, but we do not really know.

AEA: more of the same

The new electoral ballot for the American Economics Association officers is out, and it is as usual from the old boys network:

For President-elect:
William D. Nordhaus, Yale, PhD MIT

For Vice-Presidents:
John M. Abowd, Cornell, PhD Chicago
Joseph G. Altonji, Yale, PhD Princeton
Raquel Fernandez, NYU, PhD Columbia
Paul R. Milgrom, Stanford, PhD Stanford

For Executive Committee:
Amy M. Finkelstein, MIT, PhD MIT
Jonathan D. Levin, Stanford, PhD MIT
Serena Ng, Columbia, PhD Princeton
Mark D. Watson, Princeton, PhD UCSD

Consider this: one has to point out that for once there is a candidate who got a PhD from a public university. As I mentioned before, I cannot see how these people can in any way or fashion represent the profession, as there is no one from a public university, no one from a liberal arts college, no one from government, no one from private industry, no one without tenure. One does not even have a choice for president! I have a lot of respect for William Nordhaus, but if there is no choice, this election is not about representation or democracy. It is also about credibility.

The ballot allows for write-ins, though. But without coordinated action, nothing can change here. One can attempt this, though. Hence, I suggest to write in for every position Gregory Burge, assistant professor at the University of Oklahoma and PhD from Florida State University. Make it known to the AEA that these shenanigans should stop.

Vote Gregory Burge and vote often!

On the difficulty of calculating the cost of living

Quality of life indexes are popular in the press. But they are not that easy to compute. While one can easily measure how much one has to work for, say, a loaf of bread, quality of life needs to consider a broader basket of goods. Now, you need to define that basket, which may be very different across locations (and across time if the horizon is long enough). It becomes even more difficult if some of the goods are location specific, such as housing.

John Winters points out that housing rents and house values are typically used for this kind of exercise. But house values can be very misleading, as most of the price of a house contains future services and their price, not current ones. Rents, in contrast, only contain the value of current services. An additional problem is that depending on the location the rental and sale markets may be very segregated and thick. Indeed, rentals are typically small and of lower quality. Winters compares rents and house values for US metropolitan areas and finds that they correlate well, but house values exhibit wide dispersion, making them indeed less reliable. He recommends using only rents, even if few rents are available and may not be necessarily representative of the housing stock and market.