A bunch of research findings on minimum wages in New Zealand

In a new Motu Working Paper, Dave Maré and Dean Hyslop review and update a bunch of research findings in relation to minimum wages in New Zealand. There are too many results for me to adequately summarise, so I'll just focus on some of the more interesting (to me) results, and I encourage you to read the full paper if you want more.

First, over the last twenty years, New Zealand's minimum wage has increased dramatically, and in relative terms is now among the highest in the OECD:

Comparative OECD data shows NZ’s minimum wage increased from 50% of the median wage in 2000 to 59% in 2008 (and from 45% to 51% of the average wage) and was then stable at 59-60% of the median wage until 2017 (51-52% of the average wage). In comparison to other countries, these relative changes have been large with New Zealand’s minimum wage moving from about the middle of the range of developed countries to near the top...

And because the youth minimum wage was abolished, the increases in the minimum wage have been even larger for the youth population. Those changes make New Zealand an interesting case study internationally for what happens when you have very high minimum wages.

Second, there are differences in the measured rate of increase in real terms, depending on which deflator is used. As Maré and Hyslop write:

...the wedge been the CPI and PPII changes over the period which underly these differences, means that the spending power of minimum wage workers increased 75% over the period while their cost relative to firms’ other inputs only increased 55%. In this sense, the minimum wage increases have been relatively more beneficial to workers’ (consumer) spending power than they have been detrimental to firms’ production costs.

That's interesting, and a small but important piece of evidence of employers' bargaining power in the labour market, relative to workers. Higher minimum wages offset employers' bargaining power, so workers benefit more from higher minimum wages than what is costs employers.

Third, the minimum wage is most salient for younger workers, and less educated workers:

A clearer view of which groups are most strongly affected by the minimum wage is evident in... the proportion of employees in each group who are paid at or below the minimum wage. Overall, this proportion has risen from 4 percent in 1997 to 9 percent in 2020, reflecting the increase in the real value of the minimum wage. For 16 to 17 year old workers, however, around half were on the minimum wage in 2020, down slightly from 64 percent in 2008, when the youth minimum wage was abolished. Similarly, 38 percent of 18-19 year old employees were paid at or below the minimum wage in 2020. Other groups that have a disproportionate share of employees on the minimum wage include young adults aged 20-24 (20 percent), people with school qualifications as their highest qualification (15%), or with no qualification (13%), and Pasifika (13%).

There is probably an argument for looking at these groups jointly, as I suspect that the Pasifika group is also younger and less educated on average than other ethnic groups.

Fourth, Maré and Hyslop use a variety of methods and measures of the 'bite' of the minimum wage and look at the effect on employment. Briefly, they find that:

...the results presented here using either measure are comparatively underwhelming in terms of finding robust evidence of adverse employment effects associated with minimum wage increases.

That's a second piece of evidence of employers bargaining power - a lack of disemployment effects of higher minimum wages. However, it is worth noting that the time series approach isn't as robust for finding effects as other evidence internationally where the counterfactual is more obvious.

Fifth and finally, Maré and Hyslop investigate minimum wages as a form of income support, and conclude that:

...although minimum wage workers are more concentrated among lower income individuals, and to a lesser extent more likely to be in the lower income households, the latter relationship is relatively weak. Thus, as a redistributive income support policy, the minimum wage is relatively blunt and far less targeted and effective than the current set of welfare and transfer policies.

As I noted above, there's a lot of detail in the paper, and it is well worth a read if you are interested in minimum wages in New Zealand. The lack of evidence of disemployment effects of the minimum wages rises so far is potentially important. However, that should not be taken to mean that the minimum wage can be raised without limit with no effect on employment. And, if you believe that the goal of minimum wages is as a redistributive tool to reduce inequality, there is little evidence that it successfully serves that purpose. There are other policy options (starting with progressive taxation) that would be much more effective.

[HT: Eric Crampton at Offsetting Behaviour]

Security guards and bank robbery displacement

Crime is an interesting topic of study. One of the interesting aspects of it is that when a deterrent to crime is put in place, crime may be reduced in the vicinity of the deterrent, but may increase elsewhere. For example, installing CCTV cameras on main streets may reduce night-time street robberies in the areas of the CCTV cameras, but only because the criminals relocate their activities to nearby streets where there are no CCTV cameras, leaving overall crime unchanged. That is referred to as 'displacement' of crime.

In an interesting application of this theory, a new article by Vikram Maheshri (University of Houston) and Giovanni Mastrobuoni (University of Torino) looked at the effect of the location of security guards on bank robberies in Italy. They use complete data on all registered banks in Italy over the period from 2000 to 2009, which includes details of 37 different security measures. They focus their attention on the hiring (or firing) of security guards, and how that affects the number of bank robberies at the hiring (or firing) bank, and at neighbouring banks (for a variety of different neighbourhood sizes ranging from 500m x 500m grid squares to 50km x 50 km grid squares).

Using this dataset, they find that:

In markets smaller than 1 km2, we find displacement effects of 1.5 to 2 percentage points to unguarded banks... Specifically, if an unguarded bank’s neighbors hires guards, the branch’s probability of being robbed will increase by roughly 20%. However, we find no statistically significant displacement effects to guarded banks... even in the smallest markets.

No surprises there. If a criminal is thinking about robbing a bank, they may prefer to find an unguarded nearby bank when they find that their original target is guarded. However, that creates a bit of a dilemma for a policymaker (or an owner of many bank branches). Should more security guards be hired, or fewer? Maheshri and Mastrobuoni simulate the effect of different policies requiring or prohibiting security guards. They show that:

In much of the country, banning guards would lead to no more than five additional robberies. However, in metropolitan areas, we might find much greater increases. For instance, Rome, Naples, Milan, and Palermo would experience more than fifty additional robberies...

If instead all banks were required by law to hire guards... the greatest reductions in robberies are concentrated in the most densely populated areas that feature the greatest number of potential targets. These include the relatively wealthy Po’ River valley in the north (which includes Milan, Turin, Bologna, and Venice) along with the major cities of Rome, Naples, Bari, and Florence...

Hiring a security guard is costly. A rational individual bank should only do so if the cost of the guard is outweighed by the benefits in terms of bank robberies prevented (measured by the losses from bank robberies that would be avoided). That is more likely in urban areas, where banks are larger and hold more cash. However, since hiring a security guard displaces bank robberies to nearby banks, an urban bank that hires a security guard creates a problem for its neighbour banks without guards. If some (many) banks have security guards, then it is better for all urban banks to have guards. So, requiring security guards in dense urban areas makes sense. On the other hand, security guards in sparsely populated rural areas make little sense, because they don't prevent many robberies and the cost of hiring them therefore outweighs the benefits. Since few banks will have security guards, then it is better for no rural banks to have guards.

The problem with this analysis is that it is based on observations of past behaviour, and that criminal behaviour may change in response to policy changes by the banks. In the data that Maheshri and Mastrobuoni use, most of the displacement in bank robberies is very local. However, if bank robbers recognise that all urban banks suddenly have security guards, but no rural banks do, then the displacement may suddenly shift from within local neighbourhoods to between urban to rural areas. This would lead to a much wider geographical displacement effect than has been observed in the past, and place the rural banks at greater risk.

[HT: Marginal Revolution, last year]

Read more:

• The optimising behaviour of Italian bank robbers

Wikipedia could be good for research

Wikipedia may be the most maligned resource of all among academics. It has a reputation for inaccuracy that was earned during its early days, and that reputation continues to encourage scepticism over its value as a research tool. Despite that reputation and scepticism, it is clearly one of the most widely used tools by students, if not by faculty. And the quality and comprehensiveness of articles have improved greatly over time. So, I was interested to read this 2017 paper by Neil Thompson (MIT) and Douglas Hanley (University of Pittsburgh), which looks at the impact of Wikipedia on research.

Thompson and Hanley did two things. First, they used a 'big data' approach, looking at semantic word usage on Wikipedia, and whether subsequent word usage in academic articles changed to match the Wikipedia entries. Specifically, they used:

...a full edit-history of Wikipedia (20 terabytes) and full-text versions of every article from 1995 onward from more than 5,000 Elsevier academic journals (0.6 terabytes). This allows us to look at the addition of any Wikipedia article and to ask if afterwards the prose in the scientific literature echoes the Wikipedia article’s. The advantage of this approach is that we can look very broadly across Wikipedia articles.

Thompson and Hanley look at two fields: (1) chemistry; and (2) econometrics. They quickly report that the number of reads of econometrics articles on Wikipedia is far too small to generate significant effects, and bury those results in an appendix to the paper (but if you look at them, they are broadly consistent with the results from chemistry). They focus their main conclusions on the chemistry results, looking at a six-month window before, and a six-month window after, the creation of a Wikipedia article. They find that:

The positive and highly statistically significant coefficient on “After” in the regression confirms that articles published afterwards are indeed more similar.

The regression estimates are quite meaningless without more context, so they compare the effect of a Wikipedia article with the effect of a review article published in a journal, and find that:

...a Wikipedia article’s effect is roughly half as large as that of a review article. 

That's a reasonably large effect. However, their big data approach demonstrates correlation though, not causation. So, in the second part of the paper, Thompson and Hanley report on a randomised controlled trial, where they created new Wikipedia articles, and tested their effects on subsequent word usage in academic articles. As they describe:

To establish the causal impact of Wikipedia, we performed an experiment. We commissioned subject matter experts to create new Wikipedia articles on scientific topics not covered in Wikipedia. These newly-created articles were randomized, with half being added to Wikipedia and half being held back as a control group... If Wikipedia shapes the scientific literature, then the “treatment” articles should have a bigger effect on the scientific literature than the “control” articles.

Comparing their treatment and control articles, they find that:

...the scientific content from the articles we upload to Wikipedia makes its way into the scientific literature more than the content from control articles that we don’t upload...and these effects are large.

How large? They provide a back-of-the-envelope calculation that implies that:

...41 Elsevier journal articles in Chemistry were affected. If we then scale this up to account for Elsevier’s share of the journal market... then we would estimate each Wikipedia article is influencing ~ 250 scientific articles (to some extent).

They then go on to explore the results in more detail, by looking at which sections of academic articles are affected, and report that:

...there is a statistically significant effect in all sections except the abstract. The size and statistical significance is weakest in the Methods section and strongest in the Introduction. This suggests that our Wikipedia articles are having their largest effect on the contextualization of science and the connections that the authors are making to the rest of the field.

Finally, and interestingly, they find that the academic literature that is referenced on Wikipedia earns 91% more citations. This last point suggests that academics should be more active as editors on Wikipedia, not least to get their own articles referenced there and increase their citation count!

Overall, the results suggest that Wikipedia has a significant effect on the way that the academic literature is synthesised and interpreted, similar to the effect that review articles have. If that is the case, then it again suggests that academics can have a positive influence on their discipline by ensuring that Wikipedia articles accurately reflect the current state of knowledge.

[HT: Ethan Mollick on Twitter, via Marginal Revolution]

Please say no to the Auckland cycle bridge

I've finally extricated myself from the end-of-trimester marking load. While I was buried in exam and final test papers, this happened, as reported by the New Zealand Herald:

A planned standalone cycle and walkway bridge next to the Auckland Harbour Bridge has come under fire for its cost, and a newly released figure shows the cost of it could far outweigh the benefits.

Transport Minister Michael Wood did not provide the benefit-to-cost ratio (BCR) when he announced the new $785m project earlier this month – but has now revealed the initial assessment by Waka Kotahi is only 0.4 to 0.6.

That meant for every dollar spent on the bridge, there would effectively be a 40 to 60 cent loss.

Governments do some pretty dumb things, but essentially flushing money down the toilet by spending it on a project that doesn't even pass a simple cost-benefit test is about as bad as it gets. And what is especially galling about this is that the government is under fire for a slow coronavirus vaccine rollout, and for under-funding health care and education. Any of those issues could use a share of $685 million and improve wellbeing vastly more than a vanity project stretching over the Waitematā Harbour. Although, given the continuing failures on mental health despite announced additional funding, it's becoming abundantly clear that we have a government that is big on announcements and short on delivery.

I've given the government their dues over the last couple of weeks for avoiding the worst potential decisions that were open to them (see here and here), but funding this bridge is simply stupid. This priceless quote from ACT leader David Seymour pretty much sums it up:

"We'd lose less if Michael Wood sent taxpayers' money to a Nigerian prince to keep safe until he can pay us back."

Indeed.

Robert McCullough on the causes of inequality in New Zealand

Inequality is one of the great challenges of our time. As I have noted before, there is no easy solution to the problem of inequality. That's because it isn't easy to target at its source, because the causes of inequality are numerous. So numerous, in fact, that when I summarise some of the main contributors to rising inequality since the 1980s (noting that in New Zealand, almost all of the increase in inequality happened by 1995, and inequality hasn't gotten any worse since then - see here, for example) takes up a good six Powerpoint slides in my ECONS102 class (and in quite abbreviated form).

In a recent NBR article (paywalled), Robert McCullough provides a useful summary of the key drivers in the New Zealand context:

Should a government wish to reduce inequality, then its causes first need to be identified. There are four leading suspects: globalisation, increased demand for skilled labour, higher pay for top executives and government policy on things such as taxes and minimum wages...

The first suspect, globalisation, refers to how our trade with low-wage economies can lead to the pay of unskilled workers in New Zealand being “set in Beijing”. The argument is that in a global economy, domestic wages can be influenced by overseas labour-market conditions. For example, when a Kiwi-based business can run a call-centre out of the Philippines then downward pressure is put on the pay of these types of local workers...

The second suspect behind high inequality goes by the phrase ‘skill-biased technological progress’. The argument is that there has been increasing demand in the past few decades for skilled workers, which has dramatically pushed up their wages compared with unskilled workers.

A big chunk of that demand is for people with maths and computer science skills. Leading up to 2008, those trained in mathematical finance were the top dogs in London and New York. Now they are being usurped by workers in fields such as artificial intelligence, at Google and other firms, some of whom are being paid on a par with American football stars...

A third reason for higher inequality relates to the job market for top executives. One of the world’s leading names on this topic, my former British examiner, Sir Tony Atkinson, who was invited out to New Zealand by the Clark government, wrote that, “as an English-speaking country, New Zealand chief executive salaries have been most likely affected by the internationalisation of the market for executives”...

A fourth cause of inequality is government policy. The RBNZ’s poorly judged, over-the-top, $100 billion ‘quantitative easing programme’ has caused wealth inequality to rise by increasing house prices. Tweaking the top tax rate will unlikely have much of a counter effect.

McCullough's argument is that the Labour government isn't working on reducing inequality, because they aren't combatting the sources of inequality. This argument is unfair. The sources of inequality are numerous, and not every source can be easily mitigated directly. Consider how hard it would be to try to reverse skills-biased technological change, for instance. In fact, the government could do a good job in reducing inequality without directly tackling any of its causes, by making the tax system more progressive. And arguably, that's what they have been attempting to do with the new top tax rate of 39%.

Read more:

• More progressive taxation may increase economic welfare in New Zealand
• Inequality, policy and some cause for optimism
• Is inequality increasing?

The tragedy of the clinical commons

I've been meaning to write about this topic for months, but was reminded of it only when scrolling through my open browser tabs. One of the tabs was this blog post by Derek Lowe from last October, when the first clinical trials of coronavirus vaccines were about to start reporting results:

We’re getting closer to having to deal with a number of tricky issues around the first Emergency Use Authorizations (EUAs) for coronavirus vaccines. These have never quite come up in this way before, because (for one thing) EUAs for vaccines are relatively rare events, and (for another) we’ve never had so many simultaneous vaccine trials against the same disease before.

So let’s just stipulate that Somebody (be it Pfizer, Moderna, AstraZeneca, whoever) asks for an EUA before all the other Somebodies, and that this request is granted...

We may get into a situation where an interim readout of the data show that a vaccine may well be working, but that granting an immediate EUA has a real danger of blowing the statistics for the complete trial. That is truly the worst outcome: ending up with something that might be useful, but being unable (despite all the time and money and effort) to able to say if it really is. We’ve got to avoid that...

But the patients involved in all these trials may have other ideas. Each individual that decides to leave the trial protocol may feel that loss of their own data is not enough to affect the overall result, but if enough people think that way, that result will most certainly suffer – a tragedy of the clinical commons...

In the same way that you can’t force the participants of the emergency-authorized vaccine trial to stay in it, you also can’t force the participants in the other trials not to get the newly authorized one 

The Tragedy of the Commons arises in the case of common resources - goods that are rival (one person's consumption reduces the amount of the good available for everyone else), and non-excludable (if the good is available to anyone, it is available to everyone, and you can't easily prevent people from having access to it). How does this apply to the case of vaccine clinical trials? Vaccine trial participants are rival (since one participant leaving the trial reduces the number of participants remaining in the trial), and non-excludable (for medical ethics reasons, you can't stop a trial participant from leaving the trial). As each trial participant leaves the trial, fewer participants remain in the clinical trial, and eventually so few participants would be left in the trial that the results of the trial become meaningless.

The Tragedy of the Commons happens because private incentives and social incentives differ. The private incentive for the trial participants is to get vaccinated, even if in order to do so it means leaving the clinical trial they are enrolled in. The social incentive is for trial participants to remain in their clinical trial until it is completed.

Fortunately, we didn't find ourselves in a situation where one vaccine was given Emergency Use Authorisation well in advance of the others. Clinical trial participants mostly followed through on their commitment to complete the trial, and we ended up with several coronavirus vaccines that were shown to be effective. Fortunately, the 'tragedy of the clinical commons' was avoided.

[HT: Marginal Revolution]

Game developers, Tiebout competition, and the winner's curse

Regular readers of this blog (or students I have taught in ECONS102) will know that I'm not a fan of subsidising particular industries, especially when it is the industries themselves that are crying out for the subsidies. Cue Calvin and Hobbes:

So, last month's news about the gaming development incentives in Australia was a little disturbing to me. As NBR reported (paywalled):

The Australian government’s recently announced 30% tax offset for game developers could lead to a brain drain if not matched by New Zealand, an industry body has warned.

Australia’s next federal budget will include a 30% refundable tax offset for video game development as part of the government’s National Digital Economy Package.

Australia’s Interactive Games and Entertainment Association said the break would spur the creation of new Australian game development studios, accelerate the growth of existing Australian studios, and attract blockbuster game studios to Australia, creating jobs along the way.

On this side of the Tasman, New Zealand Game Developers Association chair Chelsea Rapp said the Australian tax offset would give Australian studios a massive leg-up over their New Zealand peers.

New Zealand has an existing incentive programme for film, which has successfully attracted major projects to the country, but video games are specifically excluded from the scheme.

Rapp said the NZGDA had been campaigning for this sort of government support for nearly 10 years.

There are several interrelated issues here. First, as I note in my ECONS102 class, lobbying the government to get your firm (or your industry) a better deal is essentially a socially wasteful activity. Economists refer to it as 'rent seeking' - in this case, the New Zealand gaming development industry is seeking to gain some economic rent from the taxpayer in the form of the subsidy. It's socially wasteful to add a subsidy to an existing industry, because a subsidy leads to a deadweight loss (unless there are offsetting positive externalities, which are not apparent in this case). However, if a large enough proportion of gaming developers were really going to move to Australia, the loss of economic welfare from the lost industry could (in theory) be larger than the deadweight loss of the subsidy. That might justify a subsidy.

However, that brings us to the second issue - how much should government be willing to pay to keep a gaming development industry that is threatening to move overseas. Certainly, the government shouldn't be willing to pay more than the total economic welfare that the industry generates (or, more accurately, the economic welfare that the proportion of the industry that would otherwise move overseas generates). Paying just enough so that the gaming development industry decides to stay would be the best option. A smart gaming development industry would recognise this, and play off governments against each other, leading to Tiebout competition. Tiebout competition is a 'race to the bottom', where governments compete to offer the most generous subsidies and other support, so that the gaming developers will operate in their country. This almost ensures that the 'winning' government would end up paying out all of the potential economic welfare gains directly to the gaming development industry itself.

But it gets worse. Since no government has a perfect understanding of just how much the gaming development industry is worth, the government that 'wins' the Tiebout competition will almost certainly to be the government that most over-estimates the value of the gaming development industry. This is what is referred to as the winner's curse. Over-estimating the value of the gaming development industry will mean over-incentivising it, paying too much relative to the economic welfare that it generates.

Thankfully, at this stage the New Zealand government hasn't fallen into this trap. Gaming developers don't need to be subsidised, and we definitely don't need to be in a race to the bottom to attract them here. It's bad enough that we do this for movie production, where it already doesn't pay off.

When does a hobby farmer become a used car importer? Hopefully never...

Earlier this week, the government announced final details of its proposed 'feebate' scheme to incentivise a shift to electric vehicles. As the New Zealand Herald reported earlier this week:

Drivers who buy new cars from July 1 will be able to get taxpayer-funded rebates of almost $8700 for a new electric or plug-in hybrid car, and about $3,500 for used cars.

But those who buy petrol vehicles will cop the cost under the Government's plan announced today – from January 2022, buyers of new petrol cars will have to pay a fee of up to $5875 while those buying newly imported used cars face fees of up to $2875.

That fee would be based on emissions – for example, it would add $2,900 to the cost of a new Toyota Hilux, $1230 to a Kia Sportage, and $830 for a Nissan Navara.

Incentivising a shift to electric vehicles in this way makes little sense in the presence of a binding emissions trading scheme cap. Those issues have been well explained elsewhere (see Thomas Lumley's post at StatsChat, or Eric Crampton's post at Offsetting Behaviour, for example). Instead, let's consider a potential problem that the government has so far thankfully avoided.

Farmers and contractors are up in arms about the feebate scheme, because their work vehicles will suddenly be much more expensive. As RNZ reported yesterday:

Farmers and tradies say the government's clean car package is an unfair tax on them as no alternatives are available for their work vehicles...

Federated Farmers president Andrew Hoggard suggested allowing an exemption for selected sectors - such as farming and construction - until meaningful alternatives were available in New Zealand.

[Canterbury high country farmer Simon Guild] said that was a system the rural sector could get behind.

"That makes total sense while there is no alternative, and I challenge anyone who thinks that there are alternative vehicles we can use to come to our place and I'll take them around in a high-emitting Hilux or whatever vehicle we have available and then they can tell me if that job can be undertaken by one of the current alternatives on the market."

Dunedin builder Sacha Gray said electric vehicles were similarly not up to scratch at the moment for tradies...

Gray also supported Hoggard's proposal.

I can immediately see two problems with a system that would exempt 'work vehicles' for farmers and tradespeople from the tax on imported petrol and diesel vehicles. The first is a variation on the great Jaffa Cake controversy in the UK (see also my post on a similar topic here). Who counts as a farmer or tradesperson, able to buy an imported petrol or diesel vehicle without paying the import tax? Farmers and tradespeople, you may say. Of course, but where do you draw the line? Do people with lifestyle blocks that are large enough to count as businesses for tax purposes farmers for the purpose of avoiding this tax? How big a block of land would you need to own? Do sharemilkers count? They don't own land. What about farmhands? Then that raises similar questions about tradespeople. Do handymen count? What about landlords who repair their own properties? The feebate scheme may be stupid or unnecessary for ETS-related reasons, but at least it is sensible in avoiding the necessity for a bunch of additional regulations about who is exempt.

The second issue is, once some defined group is exempted from the import tax, what stops them from buying an imported petrol or diesel vehicle, then selling it on the local second-hand market? There's no proposal to tax petrol or diesel vehicles, other than those that are newly imported. So, second-hand cars sold locally do not attract the tax. You would need another set of rules to govern how long a tax-exempt purchaser would have to hold onto their vehicle before reselling it. But it gets worse. What would then stop a farmer (or hobby farmer with a lifestyle block) from buying a bunch of Hiluxes, parking them up in a paddock, waiting out the no-resale period, then selling them, avoiding the tax, and pocketing a nice profit? Again, the feebate scheme may be stupid or unnecessary for ETS-related reasons, but at least it doesn't need to have a bunch of secondary rules to deal with people profiting from their exemptions.

The feebate scheme may be stupid or unnecessary, but at least it is not as stupid as it could get. The last thing we need is a feebate scheme that you could drive an SUV through. Literally.

No, El Salvador hasn't just turned Bitcoin into money

One of the big news items over the last week was El Salvador's announcement that Bitcoin would become legal tender. As the New Zealand Herald reported:

Cryptocurrencies have bounced back in a big way this morning after El Salvador voted to become the first nation in the world to adopt bitcoin as legal tender.

Over the weekend, the Central American nation's president crashed a wild bitcoin conference in Miami to reveal his plans for digital currencies to a rapturous reception.

Overnight, those pland [sic] edged closer to reality after they were backed by congress...

 "The #BitcoinLaw has just been approved by a qualified majority" in the legislative assembly, President Nayib Bukele tweeted after the vote late on Tuesday (local time).

"History!" the president added.

This isn't the first time that it has been asserted that Bitcoin is money. To be classified as money by economists, Bitcoin must meet certain conditions, as I noted in this 2019 post:

To an economist, money is something that fulfils three functions, which date back to William Stanley Jevons in 1875...:

1. It is a medium of exchange - you give it up when you buy goods or services, and you can receive it when you sell goods or services.

2. It is a unit of account - you can measure the value of something using the amount of money it is worth; and

3. It is a store of value - you can keep it and it will retain its value into the future.

Anything that fulfils those three functions can be considered money. So, coins and banknotes are money because you can exchange them for goods and services, you can use them to measure the value of things, and you can store them and use their value in the future.

And as I wrote last year in relation to Bitcoin:

...Bitcoin is used to buy at least some goods and services. However, Bitcoin fails on the last two criteria. It isn't used as a unit of account - no one's quoting you prices in Bitcoin, and I bet the vending machines don't report the Bitcoin price of a can of Coke. And, Bitcoin isn't much of a store of value. It fails on both of those criteria because the value of Bitcoin is far too volatile.

What has changed now? Perhaps Bitcoin will now be used as a medium of exchange, at least in El Salvador. But unless sellers start quoting prices in Bitcoin, and debts are recorded in Bitcoin, it is still not a unit of account. And Bitcoin is still too volatile to be a store of value. This year alone a single Bitcoin has fluctuated in value from US$30,000 to $US60,000. Why would a saver want to keep their savings in something that routinely gains or loses half of its value?

Even in terms of medium of exchange, there are some issues, as John Hawkins wrote this morning in The Conversation:

Consider the provision in the new law that “all obligations in money expressed in USD, existing before the effective date of this law, may be paid in bitcoin”.

Even that is complicated. How, and by whom, will the amount of bitcoins necessary to pay a debt be determined? Will it be based on the Bitcoin price at the time the debt was incurred, or when the debt falls due?

The difference of even a few days could be significant.

If the expectation is the price of Bitcoin is going to rise, why would you want to buy things with it? Why not wait? If the expectation is the price is going to fall, why would you want to accept it? For most transactions, using US dollars will still make the most sense.

Hawkins notes that there are serious risks to El Salvador's economy as a result of this move, and that the proclaimed benefits in terms of increased GDP and facilitating international remittances are not all that clear. What is clear is that this change still doesn't make Bitcoin money. 

Read more:

• You may be able to buy Coke with Bitcoin, but that still doesn't make Bitcoin money

Economic institutions and the 'Little Divergence'

In economic history, the 'Little Divergence' refers to the divergence in income per capita between countries in Europe in the early modern period (that is, after the Middle Ages). [*] Over the period from 1300 C.E. to 1800 C.E., Holland and England overtook Spain, Italy, and Portugal in terms of income per capita. This is somewhat of a surprise, given that Spain and Portugal in particular had early access to wealth from the New World.

Past research such as this article by Daron Acemoglu, Simon Johnson (both MIT), and James Robinson (Harvard), published in the journal American Economic Review (ungated earlier version here) argues that by 1500 C.E., the political institutions differed between northern and southern Europe in such a way to drive this divergence. Specifically, they argue that Spain and Portugal had absolutist monarchies, while in England and Holland the monarchies were much more constrained. Absolutist monarchies are an example of an 'extractive institution' - an institution that excludes the majority of society from political and economic power. This contrasts with 'inclusive institutions', which incentivise the participation of all people in society. You can think of societies' institutions as existing along a continuum from extractive to inclusive, with England and Holland closer to inclusive, and Spain and Portugal closer to extractive.

However, a recent paper by Antonio Henriques (Universidade de Porto) and Nuno Palma (University of Manchester) challenges this simple narrative (see also the summary of the article here). Using a variety of datasets, Henriques and Palma show that institutions were no more extractive in Spain and Portugal than in England until well after the Little Divergence had already begun.

First, looking at the number of times that parliaments met in each of the countries, they find that:

...Portugal and Spain did not perform worse than England until the mid-seventeenth century.

Second, looking at depreciation of coinage over time (which would be an indicator of an extractive ruler), they find that:

During the sixteenth century, both England and the Dutch Republic perform significantly worse than Spain and Portugal according to this measure, and it is not until the seventeenth century that the Spainish [sic] monarchy started to perform badly.

Third, they look at real interest rates in the three countries (where higher real interest rates would be an indicator that lending to the government was seen as more risky, as would be expected in a society with extractive institutions), and find that:

England and the Netherlands were not paying lower rates than their Southern rivals until late in the early modern period.

None of this is consistent with the idea that differences in institutions by 1500 C.E. were drivers of the Little Divergence. Henriques and Palma conclude that:

Iberia's economic divergence was not a consequence of inferior initial institutions. At least prior to the civil wars of the mid seventeenth century, England did not have more constraints on executive power or an environment more protective of property rights than Spain or Portugal... At some point England did have better institutions, but that point occured [sic] considerably later than 1500. Accordingly, explanations for the little divergence among Atlantic traders which rely on variation in the quality of \initial institutions" (in particular, constraints on executive power by 1500 or earlier), such as that by Acemoglu et al. (2005), cannot be correct...

Modern economic and political institutions trace their origins back to the early modern period. A lot of economic research takes advantage of this 'path dependence'. However, this paper by Henriques and Palma demonstrates the limits of path dependence as an explanation. At some points, there are going to be contemporaneous factors that influence economic and political institutions as well, and if we are not careful, we can miss those important determinants of change.

****

[*] In contrast, the 'Great Divergence' refers to the great growth in European income per capita, relative to income per capita in China and other Asian countries.

[HT: Marginal Revolution, last year]

The availability of UberX, and alcohol consumption

One of the key implications of rational behaviour is that people respond to incentives. When the cost of something decreases, we tend to do more of it. Costs are not limited to monetary costs, or even to costs that might be measured in monetary terms. Consider alcohol consumption at a bar. The 'full cost' of alcohol consumption at the bar includes the price of the alcohol purchased, and the monetary cost of travelling to and from the bar, but also the increased risk of bodily harm, and the increased risk of doing something stupid that the drinker may regret later. If any of those costs decrease, people would tend to drink more.

Now consider the introduction of cheaper travel options, like Uber. This lowers the travel cost for the drinker to and from the bar. Drinkers would tend to drink more in response to this. That is the hypothesis that this recent article by Keith Teltser (Georgia State University), Conor Lennon (University of Louisville), and Jacob Burgdorf (US Department of Justice), published in the Journal of Health Economics (ungated earlier version here), seeks to investigate.

Teltser et al. look at the introduction of UberX into US cities, and how that is related to alcohol consumption behaviour. They collect data on UberX entry and exit dates for different US cities, and use data from the CDC's Behavioral Risk Factor Surveillance System Selected Metropolitan/Micropolitan Area Risk Trends (BRFSS SMART) over the period from 2009 to 2017, for those aged 21 to 64 years. Using a difference-in-differences regression approach, they find that:

...UberX is associated with a 3.6% increase in the average number of drinks per drinking day, a 2.7% increase in drinking days, and a 5.4% increase in total drinks... We also find a 4.3% increase in the maximum number of drinks in a single drinking occasion and a 0.8 percentage point (1.33%) increase in the number of people who report any alcohol consumption over the previous 30 days.

By themselves though, those results are not definitively causal. However, Teltser et al. go on to show that:

...Uber’s impact is larger among individuals aged 21-34, including a 7.4% increase in total drinks, a 9.5% increase in binge drinking instances, and a 1.5 percentage point (2.3%) increase in the number who report any alcohol consumption in the past 30 days. Further, we find that UberX does not appear to increase drinking among those aged 65 or older.

Since younger people were more likely to have a smartphone (and more likely to have the Uber app) than older people (especially considering this was the period from 2009 to 2017), that increases the plausibility that UberX is causing more alcohol consumption. Interestingly, they also find that:

...the effects on drinking are larger in areas where transit is weaker, which reinforces the idea that Uber is facilitating away-from-home alcohol consumption.

Public transport is a substitute for UberX, so where public transport is already highly accessible, you'd expect UberX to have a smaller effect, and that is what they find. They also find that UberX is associated with increases in employment and total employee earnings at bars, but much smaller effects at restaurants (so, they aren't simply picking up some general increase in night-time entertainment).

Decreasing the full cost of drinking at bars, by making it less expensive to travel to and from bars, increases alcohol consumption. That is likely to be associated with increased harm, including harm to health from over-intoxication, or violence. However, before we conclude that UberX has an overall negative effect, we need to consider that the availability of UberX might also offset some harm by reducing the incidence of drunk driving. Teltser et al. didn't look at the effects of UberX on harm, or on drink driving, which is something we would need more information on before drawing a firm conclusion.

Don't bet on horses with fast-sounding names

In a new paper to be published shortly in the Journal of Behavioral and Experimental Economics (open access), Oliver Merz, Raphael Flepp, and Egon Franck (all University of Zurich) undertake an interesting test of behavioural economics. The affect heuristic suggests that our decisions are influenced by our emotions. In other words, the affect heuristic forms part of our System 1 thinking (to borrow from Nobel Prize winner Daniel Kahneman's excellent book Thinking, Fast and Slow), where our decision-making is fast, instinctive, and emotional.

However, the affect heuristic (and System 1 thinking more generally) is completely at odds with the Efficient Markets Hypothesis, which in its strongest form suggests that all relevant public and private information is incorporated into the price of an asset (such as a share price). The efficient markets hypothesis essentially suggests that in financial decisions we rely on System 2 thinking - slow, deliberative, and logical, and not influenced by our emotions.

Merz et al. look at the case of betting on horse racing, and in particular they look at whether the names of horses affect betting behaviour. The name of a horse should be reasonably uninformative about how fast the horse can run - there's no rule or law that says an owner can't name their slow nag Rocket Roger, for example. So, if betting behaviour is affected by the names of the horses, then that is likely to be the affect heuristic at work.

Using Betfair data from over 400,000 horse races in the UK, Ireland, the US, South Africa, and Australia, and involving nearly four million horses, they find that:

...a fast-sounding horse name has predictive power with regard to the race outcome beyond the winning probabilities implied in the odds. In particular, our results show that the winning probabilities of bets on horses with fast-sounding names are overstated, implying that the prices in betting exchange markets are not completely efficient, as prices become distorted by incorporating affective, misleading information from a horse’s fast-sounding name.

In other words, bettors place bets on horses with fast-sounding names far more often than the horses' underlying win probabilities suggest that bettors should. But before you get carried away and rush off to bet against horses with fast-sounding names:

...we find significantly lower returns for horses classified as fast-sounding compared to other horses. A simple trading strategy of betting against all horses classified as fast-sounding yields a return of approximately 2.9% before the commission but a negative return of -1.6% after deducting the standard commission of 5% from Betfair.

Even though in theory you can profit from other bettors' irrational preference for betting on horses with fast-sounding names (in apparent violation of the Efficient Markets Hypothesis), when you take into account that Betfair takes a commission on every bet, there isn't a positive profit-making opportunity here. The best you can probably do is to avoid betting on horses with fast-sounding names, because the return is going to be significantly more negative than it should be.

[HT: Marginal Revolution]

Consumer surplus illustrated

In economics, the concept of producer surplus is relatively intuitive to understand. It is the difference between what the seller receives for selling the good or service, and the seller's costs. That is effectively their profit (or, to be more correct, it's their profit if you ignore any fixed costs).

The concept of consumer surplus has a similar interpretation conceptually, but intuitively consumer surplus is more difficult to understand. It is the difference between the maximum that the buyer is willing to pay for the good or service, and the price that they actually pay.

Here's an illustration. After a tragic accident delaminated the sole of my shoe yesterday, I find myself in need of a new pair of sneakers. I went to The Warehouse this morning, and picked out a nice new pair, with a stated price of $35. Having satisfied myself that this was a good buy, I headed for the checkout.

When I arrived at the (self-service) checkout and scanned the barcode, the price came up as $19.98. Even better! I was happy to buy the shoes for $35, but now I was getting them for $19.98 instead. I have an extra $15.02 left in my bank account after the purchase than what I had anticipated. That $15.02 is consumer surplus. [*] You can think of it as a sort of profit that the buyer receives.

Consumer surplus is a measure of the benefit that buyers receive from participating in the market. If the consumer chooses not to buy, then there is no surplus - they will only buy if the price is not higher than the maximum they are willing to pay, and that means that some consumer surplus will be generated from every willing transaction between a buyer and a seller.

When we add the consumer surplus and producer surplus (and sometimes some other positive welfare effects of markets) together, we get a measure of total welfare (or total surplus). Total welfare is a concept that long-term readers of this blog will have seen many times (most recently here). It is important for understanding the value that markets generate, and how policies or interventions in the market will affect the wellbeing of the buyers and sellers operating in the market.

*****

[*] Actually, my consumer surplus is likely to be even higher than $15.02. Since I was willing to pay at least $35 for the pair of sneakers, I was probably willing to pay more. I won't tell you how much more, just in case The Warehouse is listening. We're not at the stage where retailers are extensively using personalised pricing (see here for a related example), but I'm not taking any chances!

Your electronic devices are going to cost more

John Hopkins (Swinburne Institute of Technology) wrote in The Conversation earlier this week:

The manufacturing world is facing one of its greatest challenges in years — a global shortage of semiconductors — and there doesn’t appear to be an end in sight any time soon.

According to Acer, one of the world’s largest laptop manufacturers, companies will still be affected by this shortage until at least the first half of 2022.

Semiconductors are an essential component of electronic devices, found in everything from cars and factory machinery to dishwashers and mobile phones. They harness the conducting properties of semiconductor materials (such as silicon), through the use of electric or magnetic fields, light, heat or mechanical deformation, to control the electric current flowing into a device.

Hopkins also outlines how we ended up with the shortage:

Like many current global challenges, this shortage initially began as a result of the COVID pandemic. Staff at semiconductor foundries in China and around the world were unable to go to work, plants were closed and production halted, which led to a lack of supply. The movement of that supply was also slowed down by tighter restrictions at ports and international borders.

At the same time, employees started working from home, children and students started studying from home, and many of us were confined to our homes for long periods. New equipment was needed to support these changes, establish makeshift offices and classrooms in our homes, and upgrade our existing home entertainment options. This prompted a sudden increase in demand for many of the devices that rely on semiconductors. 

These two changes (a decrease in supply, and an increase in demand) have led to the shortage of semiconductors, but the shortage will be temporary. To see why, consider the market for semiconductors as shown in the diagram below. The market was initially in equilibrium before the pandemic, with a price of PA and QA semiconductors traded. Then, during the pandemic, demand increased from DA to DB, while supply decreased from SA to SB, but the price of semiconductors initially stays at PA. At that low price, the quantity of semiconductors demanded in QD, but the supply is only QS - there is a shortage.

What happens next is important. The shortage means that some semiconductor buyers are missing out. To avoid this situation, the buyers can try to find a seller, and offer a slightly higher price to avoid missing out on some of the limited supply of semiconductors that are available. [*] In other words, the buyers bid the price upwards, and this process continues until the price reaches the new equilibrium price of PB, where there is no shortage and QB semiconductors are traded. [**]

Now, consider how the increase in the price of semiconductors will affect the market for electronic devices. Everything from phones to laptops to cars includes semiconductors, so an increase in the price of semiconductors increases the costs of production of electronic devices. This is shown in the diagram below. The price of devices is initially in equilibrium at P0, with Q0 devices traded. The increase in production costs causes supply to decrease from S0 to S1. The equilibrium price of devices increases to P1, where Q1 devices are traded. [***]

The global semiconductor shortage might seem an interesting by-product of the coronavirus pandemic. But, it is going to show up in the price of your next smartphone, laptop, or new car purchase.

*****

[*] In reality, the mechanism probably isn't quite this simple as there will be existing contracts for supply, etc. However, the buyers who least want to wait will find ways of accelerating their access to semiconductors, and that will involve offering a higher price in some way.

[**] As I have drawn it, QB is greater than QA, suggesting that the quantity of semiconductors increases. However, that need not be the case. The change in quantity is ambiguous. If the decrease in demand was a little bit smaller, the equilibrium quantity would have decreased, or could have stayed at the original quantity of QA.

[***] For simplicity, I haven't shown the increase in demand for devices here. However, that also leads to an increase in the price, meaning that prices of devices will increase even further.

Incentivising coronavirus vaccination, part 2

A couple of weeks ago, I posted about incentives for coronavirus vaccination, noting that:

...the government could find some other way to incentivise more production and consumption. Right now, we're in a situation where governments want to roll out coronavirus vaccines in the face of a substantial amount of vaccine hesitancy. Some governments have started to incentivise vaccines through more than just subsidising them and making them available for free.

Continuing the theme, the New Zealand Herald reported last week:

A Florida music event promoted will charge a premium to those who have not had the Covid-19 vaccine.

The concert promoter has come up with a discounted ticket at $18 for those who have been vaccinated, and will charge $999.99 to those who haven't.

The idea is to encourage people to get the coronavirus vaccine, especially before attending a large gathering such as a music gig...

[Leadfood Promotions' Paul Williams] said feedback for his idea has been "overwhelmingly positive" and, so far, no one has bought any of the $1000 tickets.

A general point about incentives is that, when the benefits of doing something increase, we are more likely to do it. If the benefits of vaccination now include cheaper concert tickets, that may encourage a few additional people to get vaccinated.

My earlier post about incentivising vaccinations drew on behavioural economics (specifically, prospect theory). On a related note, Alex Tabarrok also raised a good behavioural economics point recently:

A vaccination is all about immediate costs and future benefits and it’s more difficult to act on future benefits than immediate costs, ala hyperbolic discounting. A free beer, donut, or lottery ticket provides an immediate benefit to offset the immediate cost and so may encourage vaccination, especially for those who are very present oriented. Note, however, that a lottery ticket might be expected to be less beneficial than an equivalent-cost donut because the donut is truly immediate while the lottery ticket is not. On the other hand, if vaccine hesitancy is driven by over-estimated fear of rare side-effects then perhaps a lottery ticket balances with an over-estimated hope of rare-benefits.

And then this point on funding public goods through lotteries (an excellent point that I had not considered before):

Even within a risk-neutral, rational model, however, there are good reasons to tie public goods to lotteries (ungated). Charities, for example, often use lotteries or raffles to fund public goods. Why? The reason is that a lottery is a natural counter to free-riding. Imagine that there is a public good but no one contributes because they each hope to free ride off other people’s contributions. As a result, the public good is not provided. Now introduce a fixed prize lottery. If no one else contributes then a contributor wins the lottery for certain so it can’t be an equilibrium for everyone to free ride (reminiscent of my dominant assurance contract mechanism for producing public goods). Note that the lottery in this model can’t just be a regular lottery ticket where you have to match X numbers to win. It has to be a raffle where the probability of winning is 1/N where N is the number of contributors. Thus, the Maryland and Ohio vaccine lotteries, which draw winners from the vaccinated, are much better than New York version which just hands out free lottery tickets. Thus, I expect the Ohio and Maryland versions to be more successful than the New York version.

So, let me reiterate: New Zealand should be considering what incentives to put in place to encourage vaccination now. The last thing we need is to fall short of herd immunity targets, and have to continue an isolationist stance on migration, while the rest of the world is opening up.