Fair trade, coffee prices, and farmer incomes

This is my second post on Fair Trade (see my earlier post here). Having established in the previous post that sustainability labels matter for consumers (at least according to the study I reviewed in that post), the question is whether Fair Trade matters for farmers. Does it increase the prices they receive by enough to offset the certification costs? Does it increase farmer incomes?

The basic mechanics of the Fair Trade pricing system work like this: farmers receive a stated minimum price for coffee or the market price, whichever is the greater. They also receive an additional premium, which is to be used only for social or business development purposes, as democratically determined by members of each cooperative.

A recent paper (ungated longer and earlier version here) by Alain de Janvry (UC Berkeley), Craig McIntosh (UC San Diego), and Elisabeth Sadoulet (UC Berkeley) argues that easy (not quite free) entry into the supply of fair trade coffee eliminates any premium that coffee producers receive from being Fair Trade certified. Since the barriers to entry into Fair Trade production are relatively low, if the Fair Trade premium is high many producers seek to become certified, competing with the existing suppliers for a limited market for Fair Trade coffee, and reducing the proportion of each supplier's coffee that is sold at the Fair Trade premium (effectively reducing the average price received).

This outcome is driven by the actions of the certifiers. Because the certifiers only source of income is to provide certification, the incentive is for them to over-certify relative to the number of certified farmer cooperatives that would maximise farmer profits. Since the farmers pay for the certification, de Janvry et al. argue that the result is:

...that the price premiums in the FT system have largely flowed toward certifiers rather than producers as intended by the consumers of FT coffee.

They demonstrate their results using data from Coffeecoop, a Central American association of coffee cooperatives where all coffee is Fair Trade certified, over the period 1997 to 2009. However, because not all Fair Trade certified coffee is sold as Fair Trade, the authors can compare Fair Trade prices with non-Fair-Trade prices for the same batch of coffee, which allows them to control for quality (which is an important determinant of coffee prices).

They show that:

...the nominal premium was quite significant in the years 2001 to 2004 with low NYC price, reaching an average of 60c to 64c per pound over a market price of 63c per pound but falling to 5c to 9c per pound over a market price of 126c per pound from 2006 to 2008, even though the social premium in these years should have been at least 10c per pound.

That seems quite good overall, until you consider what happens to the proportion of coffee sold as Fair Trade, where they confirm that the sales share moves inversely with the Fair Trade premium (i.e. when the premium is high, the cooperative is able to sell a lower share of coffee as Fair Trade). The overall result is that:

The average effective premium over the thirteen years of our data is 4.4c per pound over an average NYC price of 107c per pound, and only 1.8c per pound over the last five years, 2005 to 2009.

And when they consider the costs of certification, which they estimate at 3c per pound:

...the average result of participating in the FT market has been only 2.5c per pound for the whole period of observation and with a loss of 1.2c per pound over the last five years.

You may consider that as only one result based on a single case study, and while it looks at prices at the cooperative level, it doesn't look at the effects at the farmer's level.

Ana Dammert and Sarah Mohan (both Carleton University) recently reviewed the literature on the economics of Fair Trade in the Journal of Economic Surveys (ungated earlier version here). They point out that establishing the impacts of Fair Trade for farmers is not straightforward - there is likely to be a selection bias, since farmers who expect to do well out of certification are more likely to choose to become certified. Few studies have accounted for this well, and those that have done have mostly used some form of propensity-score matching, where they essentially compare farmers that are certified with those who aren't but otherwise look very similar to the certified farmers (e.g. same size farm, same land quality, same education, same farm assets, etc.).

In terms of these higher-quality studies, the one study using this approach to look at prices found no evidence of increases in prices received by farmers. And for incomes, there are some gains but those gains are small relative to alternative income generation activities for rural dwellers like migration or employment in the rural non-farm economy. Having said that, they also find that:

Another strand of the literature accounts for selection bias and shows that Fair Trade producers have better assets, higher rates of savings and higher levels of animal stocks and perceive their land as having high renting value...

That is difficult to reconcile with little increase in incomes. So, it's possible that there are small gains for coffee farmers from Fair Trade certification. However, it's hard to say that those gains justify the much higher prices that coffee consumers pay (although the consumers do receive a 'warm glow' benefit as well).

Finally, maybe some of the other labelling initiatives are better for farmers than Fair Trade? Consumers were willing to pay more for Rainforest Alliance labelling than Fair Trade, after all. I'll return to this last point in a future post.

Read more:

• Sustainability labels do matter

Douglass C. North, 1920-2015

I'm a bit late to this due to Thanksgiving-related activities here in the U.S., but Nobel laureate Douglass C. North passed away earlier this week. North shared the 1993 Nobel Prize with Robert Fogel (who passed away in 2013) for "having renewed research in economic history by applying economic theory and quantitative methods in order to explain economic and institutional change". North's work led to the development of both cliometrics (the quantitative study of economic history) and new institutional economics.

I use a little bit of North's work in my ECON110 class, where we spend half a topic on property rights and their historical development in western countries. He is also one of the co-authors of the required textbook for that class, which is up to its 19th edition.

Washington University in St Louis has an obituary here, and the New York Times also has an excellent obituary. Tyler Cowen has collected a number of links on Douglass North here.

It is really sad - we seem to be going through a bad few years in terms of the loss of economics Nobel winners.

Sustainability labels do matter

I've been reading a few papers on aspects of Fair Trade recently (I'll blog some of them over the coming days). Like many economists, I'm not sold on the positive effects of fair trade - but more on that later. In this post, I want to focus on this recent paper by Ellen Van Loo (Ghent University), Vincenzina Caputo (Korea University), Rodolfo Nayga Jr. and Han-Seok Seo (both University of Arkansas), and Wim Berbeke (Norwegian Institute for Bioeconomy Research), published in the journal Ecological Economics (sorry I don't see an ungated version anywhere).

In the paper, the authors do a couple of interesting things. First and foremost, they use discrete choice modelling (a type of non-market valuation technique, where you repeatedly present people with different hypothetical options, and they choose the option they prefer - a technique I've used for example in this paper (ungated earlier version here), and in a forthcoming paper in the journal AIDS and Behavior that I'll blog about later), to investigate people's willingness-to-pay (WTP) for different sustainability labels on coffee. The different labels they look at are USDA Organic certified, Fair Trade certified, country-of-origin labelling, and Rainforest Alliance certified. If people truly value these difference labels (presumably because of the certification), then they should be willing to pay more for products that have them.

Second, the authors use eye-tracking technology to follow which characteristics of the products people pay the most attention to when making these hypothetical choices. Eye-tracking involves following the movement of the eyes so that you can identify where the subject is looking, and for how long they are concentrating on elements they are looking at. I'd say it's quite an exciting thing to do in the context of discrete choice modelling, since there is always the change that people don't pay attention to all of the attributes of the hypothetical products (or scenarios) they are presented with.

Anyway, the authors found a number of things of interest, starting with:

When evaluating the coffee attributes, participants attached the highest importance to the flavor followed by the price, type of roast and in-store promotions... the sustainability labels are perceived as less important compared to other coffee attributes, with USDA Organic and Fair Trade being more important than Rainforest Alliance.

They also discovered there were three distinct types of consumers:

1. "Indifferent" - consumers who didn't pay attention to either price or sustainability labelling (9.9% of the sample);
2. "Sustainability and price conscious" - consumers who attached a high importance to both sustainability labelling and price (58.0% of the sample); and
3. "Price conscious" - consumers who attached a high importance to price but not sustainability labelling (32.1% of the sample).

The eye-tracking results confirmed that the consumers who said they attached higher importance to sustainability labelling (or price) did indeed pay more attention to those attributes when selecting their choice in the discrete choice exercises. But did they want to pay more for them? The authors find that:

USDA Organic is the highest value attribute [among the sustainability labels]... followed by Rainforest Alliance and Fair Trade...

USDA Organic had the highest WTP among all the sustainability labels examined, resulting in a WTP premium of $1.16 for a package of 12 oz. This is followed by the Rainforest Alliance label and the Fair Trade label ($0.84 and $0.68, respectively).

So, people are willing to pay more for sustainable coffee. How does that compare with what they actually pay though? The authors note:

The actual price premium for coffee with a sustainability label ranges from $1.5 to $2.3/12 oz. when comparing coffee products with and without the label from the same brand.

Which suggests that the retailers are over-charging for these sustainable coffee products, relative to what the average consumer is willing to pay. However, the results overall do suggest that sustainability labels do matter. It doesn't tell us whether that is a good thing overall though - a point I'll no doubt come back to later.

Try this: Regional activity report

Last week I wrote a post about tourism, that looked at data from the Ministry of Business, Innovation and Employment (MBIE)'s Regional Activity Report. This online data is a treasure-trove of summary statistics for all of the regions and territorial authorities.

If you scroll down you can also see how the regions and territorial authorities compare across eight sets of indicators:

1. Social and Income - including household income, household income distribution, earnings by industry, deprivation index, and internet;
2. Housing - including mean weekly rent, median house price, mean house value, and new dwellings;
3. Workforce - including employment rate, labour force participation rate, NEET rate, unemployment rate, quarterly turnover rate, employment by industry, and employment by occupation;
4. Education - including national standards achievement, and NCEA Level 2;
5. Population - including population estimates, population projections, international migration, population by ethnicity, population by age group, and rural-urban proportions;
6. Economic - including GDP per capita, GDP by industry, businesses by employees, new building consents, and new car registrations;
7. Agriculture - including agricultural share of regional GDP, and area in farms; and
8. Tourism - including guest nights per capita, accommodation occupancy rate, tourism spend, international guest nights, and international visits.

In most cases the data tracks changes over time as well, some of it back to 2001. 

While most of this data was already freely available (from Statistics New Zealand, mostly), having it all collected in a single place and in a very user friendly interface, makes it an excellent resource. Even better, you can easily download any of the data into CSV files to play with yourself.

I won't provide an example of what you can do with it. I'm sure you're all more than capable of playing with the data yourselves. Enjoy!

Streams aren't drowning music sales

A couple of weeks back I wrote a post about digital music sales, specifically about Radiohead's short-lived free offering of their album In Rainbows, and how it didn't impact on their digital music sales. A broader question, that we discuss in ECON110, is how new business models are affecting more traditional revenue streams in the media markets. Take the example of streaming music (or streaming video), which has exploded in popularity in recent years (although not everyone believes this is necessarily an example of a new industry - witness the notorious views on Spotify of Radiohead's Thom Yorke).

The important question, from the point of view of the artists, is does streaming pay off? Does it cannibalise music sales (digital or otherwise)? A recent paper (ungated version here) by Luis Aguiar (Institute for Prospective Technological Studies in Spain) and Joel Waldfogel (University of Minnesota) looks specifically at this question.

There are third main possibilities with digital streaming. First, they might cannibalise music sales, because listeners find streaming cheaper or more convenient than owning their own catalogue of music. Second, they might stimulate music sales, perhaps because they allow people to sample music they otherwise would not have heard. Third, they might increase revenues despite having little effect on music sales, because consumers who previously downloaded the music from illegitimate sources instead decide to stream the music. Or some combination of these.

The biggest problem with looking at this econometrically is that if you simply run a regression with music sales and streaming, you are likely to observe a positive relationship, simply because more popular music tracks will attract both more music sales and more streaming. You can tell a similar story about pirating of music. Indeed, when Aguiar and Waldfogel look at song-level data, they observe positive relationships.

So, instead they move to looking at aggregated data, which will be less vulnerable to song-level popularity effects. Specifically, they use weekly data on song-level digital sales in each of 21 countries for 2012-2013, artist-level piracy via torrents, both aggregated to the country level, and a country-level index of Spotify use. They find several important results:

First, when we use aggregate data, we find sales and piracy displacement by Spotify. Second, when we use data on the US covering the period of substantial change in Spotify use, we find smaller sales displacement than when we use 2013 data. Third, the coefficients relating Spotify use to recorded music sales differ by format. The coefficients for digital track sales are generally higher than the coefficients for albums, but many of the album coefficients are also negative and significant. Our best estimate indicates than an additional 137 streams displaces one track sale.

Now, that isn't the end of the story. Artists should be less worried about the degree of cannibalisation per se than what it does to their revenue. Streaming adds to artist revenue, while lost sales obviously reduce revenue. What is happening to revenue overall? Aguiar and Waldfogel show that, based on reasonable assumptions about revenue per stream, that Spotify is essentially revenue-neutral for artists.

Finally, there are a few reasons to be wary of these results. First, as Andrew Flowers has noted, the per-stream artist revenue that Spotify claims is contested. So, perhaps it is lower and leads to net decreases in artist revenue. Second, the results are based on data from streams of top-50 songs, which are extrapolated to all streams. There is no way to know whether this extrapolation is reasonable without access to Spotify's proprietary data (and they're not volunteering it - if I was cynical, I might point out that this is awfully convenient!). Third, as Andrew Flowers points out, we don't know about the distribution of artist revenue. It's likely to be concentrated among a few high-profile artists, with the little guys getting very little. So, streaming might be net positive for the big artists and net negative for the little artists (or the other way around), and net neutral overall. Again, this is something we couldn't evaluate without more disaggregated Spotify data (and a good identification strategy to overcome the issues noted above).

[HT: Marginal Revolution]

The gender bias in economics

A fair amount has been written over the last couple of weeks about gender bias in economics. This Justin Wolfers piece in the New York Times was one of the catalysts, and it was followed up by Dan Diamond at Forbes and Jeff Guo at the Washington Post. The storm has been mostly about the new paper by Anne Case and Angus Deaton, which has more often than not been reported as a Deaton paper, with Case mentioned almost as an afterthought (so it's worth noting that Anne Case is a well-respected economics professor at Princeton in her own right).

Tyler Cowen at Marginal Revolution then pointed to this new paper by Heather Sarsons (a PhD candidate at Harvard, where I am based for the next month), entitled "Gender differences in recognition for group work" (which was also reported on by Jeff Guo in his article). In the paper, Sarsons first notes that there is a well-established literature noting that women are less likely to be promoted than men, and that "over 30% of the observed gap in tenure rates can not be accounted for by observable productivity differences or family commitments". She then looks specifically at tenure decisions for economics faculty, testing whether co-authoring of papers prior to tenure decisions has different effects for male and female academic economists.

The choice of economics as the field to study is deliberate. In many fields, the order of authorship in co-authored papers is meaningful - the first author was often the largest contributor, the second author was the second-largest contributor, and so on (how 'large' relative contributions are is open to negotiation, I guess). In contrast, in economics it is more likely that co-authors appear in alphabetical order, regardless of the merit of their relative contributions [*]. So, while in other fields the order of authors' names provides a signal of their relative contributions to co-authored papers, this typically isn't the case for economics. Sarsons's hypothesis is that this leads employers (universities) to have to make judgment calls about the relative contributions of the co-authors, and that these judgment calls tend to go against women (because the employers' prior belief is that female economists are lower quality than male economists).

Using data from 552 economists over the period 1975 to 2014, she first finds that:

Approximately 70% of the full sample received tenure at the first institution they went up for tenure at but this masks a stark difference between men and women. Only 52% of women receive tenure while 77% of men do. There is no statistically significant difference in the number of papers that men and women produce although men do tend to publish in slightly better journals...

An additional paper is associated with a 5.7% increase in the probability of receiving tenure for both men and women but a constant gender gap between promotion rates persists. Women are on average 18% less likely to receive tenure than a man, even after controlling for productivity differences. 

So, women received tenure at a lower rate than men, but why? The total number of papers they publish is no different and doesn't make a difference to the probability of tenure, and the difference in paper quality is actually rather small (even though it is statistically significant). Turning to her specific hypothesis about co-authorship, she finds that:

an additional coauthored paper for a man has the same effect on tenure as a solo-authored paper. An additional solo-authored paper is associated with a 7.3% increase in tenure probability and an additional coauthored paper is associated with an 8% increase.

For women, a sole-authored paper has an effect that is not statistically significantly different from that for men, but the effect of an additional co-authored paper is nearly 6 percentage points lower (i.e. a 2% increase, rather than an 8% increase, in the probability of receiving tenure).

To test the robustness of her findings, she does a similar analysis with sociologists (where the social norm is authorship by order of relative contributions), and finds no significant differences in tenure decisions between men and women. She concludes:

The data are not in line with a traditional model of statistical discrimination in which workers know their ability and anticipate employer discrimination...

The results are more in line with a model in which workers do not know their ability or do not anticipate employer discrimination, and where employers update on signals differently for men and women.

So, there is a bias against women. How far does this bias extend? There is a conventional wisdom that men are more suited for economics than women (to which I say: they should attend one of my ECON110 classes, where the female students are more often than not at the top of the class). This recent paper by Marianne Johnson, Denise Robson, and Sarinda Taengnoi (all from University of Wisconsin Oshkosh) presents a meta-analysis of the gender gap in economics performance at U.S. universities. Meta-analysis involves combining the results of many previous studies to generate a single (and usually more precise) estimate of the effect size. It (hopefully) overcomes the biases inherent in any single study, such as the study of gender gaps in economics performance.

In the paper, the authors [**] take results from 68 studies, containing 325 regressions. They find a number of things of note:

...only 30.7% of regressions actually conform to the conventional wisdom - that men statistically significantly outperform women... In 9.2% of regressions, we find that women performed statistically significantly better...

Notable is the increase in regressions that find that women outperforming [sic] men after 2005...

We find a negative coefficient on year, which would indicate that the performance gap is narrowing over time, regardless of whether we use the year of data collection or the year of publication as our time measure.

So, the observed performance gap is declining over time. Which brings me back to Sarsons's paper. She used data from 1975 to 2014, which is a wide span of time, over which things have gotten better for female academics (surely?). I wonder what happens if she accounts for year of tenure? I doubt the effect goes away, but at least we might know a bit more about the changes over time and if indeed things are getting better.

[Update: I just had one of those wake-up-and-realise-you-said-something-wrong moments. Sarsons does include year fixed effects in her econometric model, so is already controlling for changes over time somewhat, but not changes in the difference in tenure probability between men and women over time (which would involve interacting that fixed effect with the female dummy or with one or more of the other variables in her model)].

[HT: Shenci Tang and others]

*****

[*] There are many exceptions to this co-authoring rule. My rule-of-thumb in co-authoring is that the first author is the one whose contributions were the greatest, with all other authors listed alphabetically.

[**] All women co-authors, which relates to the Sarsons's paper, since one of her other results was that co-authoring with other women didn't result in as large a penalty as co-authoring with men.

Consumer technology and insurance fraud

This week is International Fraud Awareness Week. So, this story in the New Zealand Herald last week was a bit early. It notes:

Insurance fraud spikes whenever a new technology upgrade, such as the release of a new iPhone, occurs, says Dave Ashton, head of the Insurance Council's Insurance Claims Register...

Speaking at the council's annual conference this week, Ashton said when new technology was released, many people want to upgrade their model and therefore claimed their older models had been stolen, lost, or accidentally damaged. The ensuing insurance pay-out funded the new upgrade.

This type of insurance fraud is an example of what economists call moral hazard. Moral hazard is the tendency for someone who is imperfectly monitored to take advantage of the terms of a contract (a problem of post-contractual opportunism). Smartphone owners who are uninsured have a large financial incentive to look after their devices and avoid dropping them in the toilet (which I note one of our previous ECON100 tutors did, twice), because if they damage the device they must cover the full cost of repair ore replacement themselves (or use a damaged phone, etc.). Once their phone is insured, the driver has less financial incentive to look after their device because they have transferred part or all of the financial cost of any accident onto the insurer. The insurance contract creates a problem of moral hazard - the smartphone owner's behaviour could change after the contract is signed.

Things are worse in the case of the type of insurance fraud noted in the article. This goes beyond simply being less careful, and extends to deliberate misrepresentation to take advantage of the terms of the insurance contract.

If you want to understand why, you only need to consider the incentives. A rational (or quasi-rational) person makes decisions based on the costs and benefits, e.g. the costs and benefits of claiming that your old phone was stolen. The benefit is the added benefit from the new phone (compared with the old phone). The costs are the moral cost (from lying to the insurance company), as well as the expected cost associated with being caught (which depends on the probability of being caught, and the financial penalty you face if caught).

Most people don't engage in insurance fraud because the combined costs (moral plus potential financial punishment) are greater than the perceived benefits. However, when a new phone is released the benefits of fraud increase significantly (because the new phones provide a much increased benefit). Again, this doesn't mean that everyone will engage in insurance fraud, but at least some people will.

Insurance companies are not stupid though, and they are moving on this, according to the Herald article:

The council this week launched upgraded technology on the register which allows more 'big data' analysis, including hotspots such as the number of burglary claims in a particular neighbourhood. It can also provide predictive analysis around likely claims for things like weather events based on the collective claims history.

So you can probably expect your next claim for a damaged iPhone to come under a little more scrutiny, if you are claiming just after a new phone is released.

Read more:

• How to beat the fitness trackers (for cheaper health insurance)
• Why your Fitbit or Apple Watch could soon get you cheaper health insurance
• Could technology eliminate moral hazard in car insurance?

Uber just told us there is elastic demand for rides

The New Zealand Herald ran a story on Tuesday about Uber cutting fares by 10 per cent in New Zealand. In the story, they quote the general manager for Uber in New Zealand, Oscar Peppitt:

"When we are able to cut our prices, that actually leads to an increase in earnings for drivers...

When we're able to drop those prices, earnings for drivers increase, and increase disproportionately to what we drop."

Essentially, Peppitt is telling us that demand for rides on Uber is what economists term price elastic. When demand is elastic, that means that an increase in price will result in a more than proportional decrease in the quantity sold (e.g. a 10% increase in price might lead to a 20% decrease in quantity). The reverse is also true - a decrease in price will result in a more than proportional increase in the quantity sold (e.g. a 10% decrease in price might lead to a 20% increase in quantity).

When demand is elastic, decreasing your price will increase total revenue. This is because for a simple firm (like a taxi) total revenue is simply price multiplied by quantity. If price decreases, but quantity increases by a greater proportion, then total revenue will increase.

Here's where things get tricky, because firms don't operate in isolation. Elasticities have a strategic element, as we teach in ECON100. Uber might face elastic demand for rides when it lowers prices and other taxi firms don't match the price decrease. Taxi customers would suddenly find that Uber's rides are much cheaper than those of other taxi firms (they are already, but we won't worry about that for now). Many customers will switch to Uber, leading to a large increase in demand (relatively elastic demand). This is illustrated in the diagram below. When Uber lowers their price from P0 to P1, and no other firm matches the new price, then Uber faces the demand curve D1. Quantity increases by a lot, from Q0 to Q1. Total revenue, which is the rectangle under the price and to the left of the quantity, increases from P0*Q0 to P1*Q1 (it should be clear the rectangle of total revenue becomes larger).

However, if the other taxi firms also lower their prices, Uber doesn't have the same cost advantage. There will be some increase in customers overall (because prices are lower for customers from all firms), but Uber won't capture all of those new customers because the other firms are cheaper now too. That means that the quantity demanded from Uber will increase, but not by as much. In the diagram above, when Uber lowers their price from P0 to P1, and other firms also lower their prices, then Uber faces the demand curve D2. Quantity increases a little, from Q0 to Q2. Total revenue decreases from P0*Q0 to P1*Q2 (it should be clear the rectangle of total revenue becomes smaller).

To make matters worse, if marginal costs are increasing then not only does total revenue decrease, but profits for taxi drivers will decrease too (note that this is a possibility even if total revenue increases slightly). And we haven't even considered the distributional impacts (some Uber drivers might get lots more rides, while others get the same, depending on when they are available to drive, where they are located, and so on). So it's probably overly simplistic to assume that lower prices and more rides will make all Uber drivers better off. It certainly shouldn't attract more drivers to sign up for Uber.

So maybe Uber has other motives for lowering prices? Market penetration pricing perhaps? Or just good publicity to try and capture market share?

More on hobbits and tourism

A few weeks back, I posted about the impact of Lord of the Rings on tourism arrivals in New Zealand. The conclusion was that there was a short-term rise in tourist arrivals to New Zealand after the films, but that the effect did not persist in the longer term.

Last week the Herald ran a story about the impact of the Hobbit films on local tourism, specifically tourist spending in the Matamata-Piako District (where the Hobbiton Movie Set is located). The story was backed up by an impressive data visualisation on the new Herald Insights site. From the story:

The Hobbit film trilogy has catalysed a spending surge in the Matamata region by tourists from the likes of Australia, Germany, United Kingdom and North America.

The amount spent in the area by those tourists has risen at a greater magnitude over the past five years, relative to 2009 spending, than in any other region in New Zealand.

Of course, this is great news for the Matamata-Piako District, as it means more tourist spending, and more jobs in tourism, accommodation, and other services. However, it doesn't mean that overall tourist arrivals have increased (thankfully the Herald story doesn't imply this either), and one might rightly wonder which areas may have lost tourism spending as a result of tourists flocking to Matamata instead?

MBIE's regional activity report is an outstanding interactive tool for taking at least an initial look at these questions. Expanding on the Herald's example, German tourists' spending in Matamata-Piako increased by 535% between 2009 and 2014. The big losers (of German tourist spending) over the same period appear to be Porirua City (down 50%), Hauraki District (down 40%), and Palmerston North City (down 28%). See here for details.

It is also worth noting that German tourists were responsible for just 1.9% of tourist spending in Matamata-Piako in 2014. The trend in increased spending is apparent across many groups for Matamata-Piako - there are similar (but not as large in relative terms) spikes for spending by tourists from the rest of Europe (excluding Germany and the UK), the U.S., Canada, and Australia. But not for China or Japan.

So, an overall win for Matamata-Piako, but hard to say whether it is a net win for New Zealand.

Joseph Stiglitz on high frequency trading

I’ve had the occasional disagreement with my more left-leaning students about my views on markets. Usually, they’ve gotten the wrong impression of where I stand, which is that although markets are usually a good thing, there are limits. One of the things that has always quietly concerned me has been the way that financial markets work (or sometimes, fail to work). So, when I read this piece by Joseph Stiglitz (2001 Nobel Prize winner), I was quite happy to see that we share some disquiet in this area. Overall, Stiglitz appears to be quite negative on the effects of high frequency trading.

The paper is very readable (not uncommon based on other writing by Stiglitz I have read), even for those who are not technically inclined. Some highlights:

High frequency trading… is mostly a zero sum game—or more accurately, a negative sum game because it costs real resources to win in this game...

A market economy could not operate without prices. But that does not mean that having faster price discovery, or even more price discovery necessarily leads to a Pareto improvement… it is not obvious that more trading (e.g. flash trading) will result in the markets performing price discovery better...

…there may be little or no social value in obtaining information before someone else, though the private return can be quite high… And because the private return can exceed the social return, there will be excessive investments in the speed of acquisition of information.

…if sophisticated market players can devise algorithms that extract information from the patterns of trades, it can be profitable. But their profits come at the expense of someone else. And among those at whose expense it may come can be those who have spent resources to obtain information about the real economy… But if the returns to investing in information are reduced, the market will become less informative.

…managing volatility and its consequences diverts managerial attention from other activities, that would have led to faster and more sustainable growth.

Stiglitz concludes:

While there are no easy answers, a plausible case can be made for tapping the brakes: Less active markets can not only be safer markets, they can better serve the societal functions that they are intended to serve.

In other words, when it comes to financial markets, sometimes less is more.

[HT: Bill Cochrane]

Corporate prediction markets work well, given time

Prediction markets were all the rage in the 2000s. The Iowa Electronic Markets were at their peak, and James Surowiecki wrote the bestseller The Wisdom of Crowds. The basic idea is that the average forecast of a bunch of people is better than the forecast of most (or sometimes all) experts. I was quite surprised that prediction markets appeared to go away in recent years, or at least they weren't in the news much (probably crowded out by stories about big data). I was especially surprised we didn't see many stories about corporate prediction markets, which suggested they weren't particularly prevalent. It turns out that wasn't the case at all.

This paper in the Review of Economic Studies (ungated earlier version here) by Bo Cowgill (UC Berkeley) and Eric Zitzewitz (Dartmouth College) shows that this wasn't the case at all, as demonstrated by their Table 1:

The paper has much more of interest of course. It looks at three prediction markets, at Google, Ford, and an unnamed basic material and energy conglomerate (Firm X), and tests whether these markets are efficient. They find:

Despite large differences in market design, operation, participation, and incentives, we find that prediction market prices at our three companies are well calibrated to probabilities and improve upon alternative forecasting methods. Ford employs experts to forecast weekly vehicle sales, and we show that contemporaneous prediction market forecasts outperform the expert forecast, achieving a 25% lower mean-squared error... At both Google and Firm X market-based forecasts outperform those used in designing the securities, using market prices from the first 24 hours of trading so that we are again comparing forecasts of roughly similar vintage.

In other words, the prediction markets perform well. There are some inefficiencies though - for instance, Google's market exhibits an optimism bias, which is driven by traders who are overly optimistic about their own projects (and their friends' projects), as well as new hires being especially optimistic. However, the inefficiencies disappear over time, and:

Improvement over time is driven by two mechanisms: first, more experienced traders trade against the identified inefficiencies and earn higher returns, suggesting that traders become better calibrated with experience. Secondly, traders (of a given experience level) with higher past returns earn higher future returns, trade against identified inefficiencies, and trade more in the future. These results together suggest that traders differ in their skill levels, they learn about their ability over time, and self-selection causes the average skill level in the market to rise over time.

So, prediction markets work well for firms, given enough time for inefficiencies to be driven out of the markets. This is what you would expect - traders who are consistently poor forecasters either drop out of the market or they learn to be better forecasters.

However, as the authors note they were limited to looking at just three prediction markets, and only those who would share data with them. There is likely to be some survival bias here - prediction markets that don't work well won't last in the market long, and are unlikely to be observed. On the other hand, by the time of writing the paper, the markets at Google and Ford had closed down in spite of their good overall predictive performance. On this last point, the authors note that "decisions about the adoption of corporate prediction markets may... depend on factors other than their utility in aggregating information". Other forecasters don't like being shown up.

[HT: Marginal Revolution]

Try this: Crash course economics

There is a new video series on YouTube called Crash Course Economics. The videos star Adriene Hill and Jacob Clifford, and are well worth looking at. Here's the first installment:

Enjoy!

[HT: Mark Johnston at Econfix]

Cost-benefit analysis of averting catastrophes just got a whole lot harder

I'm a bit late to this, given it has already been covered by Tyler Cowen and then by Eric Crampton. However, now that I've read the paper in the American Economic Review (ungated earlier version here) by Ian Martin (London School of Economics) and Robert Pindyck (MIT), I feel I can comment on it.

In the paper, the authors develop a mathematical theory of the costs and benefits of averting (or reducing the impact of) catastrophes (like viral epidemics, Nuclear or bioterrorism, climate change catastrophes, etc.). The core (and somewhat worrying) takeaway from the paper overall is that, in the presence of multiple potential catastrophes, a simple cost-benefit rule doesn't work:

Conventional cost-benefit analysis can be applied directly to “marginal” projects, i.e., projects whose costs and benefits have no significant impact on the overall economy. But policies or projects to avert major catastrophes are not marginal; their costs and benefits can alter society’s aggregate consumption, and that is why they cannot be studied in isolation...

When the projects are very small relative to the economy, and if there are not too many of them, the conventional cost-benefit intuition prevails: if the projects are not mutually exclusive, we should implement any project whose benefit wi exceeds its cost τi. This intuition might apply, for example, for the construction of a dam to avert flooding in some area. Things are more interesting when projects are large relative to the economy, as might be the case for the global catastrophes mentioned above, or if they are small but large in number (so their aggregate influence is large). Large projects change total consumption and marginal utility, causing the usual intuition to break down: there is an essential interdependence among the projects that must be taken into account when formulating policy.

The implications of this are pretty broad, including:

• The value (to society) of averting a catastrophe depends on what other catastrophes are being averted;
• It may be optimal not to avert some catastophes, even when averting those catastrophes might seem justified based on a naive cost-benefit evaluation;
• Deciding which catastrophe is the most serious and prioritising averting that one is not the optimal approach; and
• These results hold if there are many small catastrophes, as well as when there are fewer (but more serious) ones.

The authors conclude:

We have shown that if society faces more than just one catastrophe (which it surely does), conventional cost-benefit analysis breaks down; if applied to each catastrophe in isolation, it can lead to policies that are far from optimal. The reason is that the costs and benefits of averting a catastrophe are not marginal, in that they have significant impacts on total consumption. This creates an interdependence among the projects that must be taken into account when formulating policy.

All of which means that, because of the interdependence between these catastrophes (even if they are many and small), cost-benefit analyses of things like mitigating earthquake or tsunami risk just got a whole lot harder to do in a robust way.

A sugar tax wouldn't just be paid by the manufacturers

There is a very common misperception in the public about who pays the cost of excise taxes (taxes on the sale of goods or services). In most respects, it actually does not matter whether a tax is levied on the sellers (producers) or buyers of the product or service - the tax will be shared between both the sellers and the buyers.

As one recent example of this misconception, see this comment by Gwen in the NZ Herald Rants and Raves yesterday:

After reading the item in the Herald about introducing a sugar tax, I think surely the manufacturers are to blame, and therefore they should be paying a sugar tax, not the taxpayers who already have plenty of tax costs. Come on manufacturers, take a sugar pill and reduce sugar in all your products. Life is so sweet.

Unfortunately, if a sugar tax was introduced there is no way to ensure that it would only affect manufacturers. To see why consider the diagram below, which represents the market for some sugary food product (for simplicity, we'll assume that the sellers in this market are the manufacturers). With no tax, the market operates in equilibrium with the price P0 and the quantity of sugary food products traded is Q0.

Now say that the government introduces a tax - we'll assume it is what we call a 'specific tax', which is a constant per-unit amount (e.g. $1 per sugary food product). Let's say the government levies this tax on the manufacturer - the manufacturer must pay this tax to the government. The effect of this is like increasing the costs of supplying the market (because in addition to their usual costs of production, the manufacturer must now pay a tax for every unit of sugary food product they produce). To represent this on our market diagram, we create a new curve (S+tax), which represents the costs of supply (S) plus the cost to the manufacturer of the tax. In effect the S+tax curve is vertically above the supply (S) curve by exactly the per-unit dollar value of the tax.

With this new tax in place, the manufacturer will increase the price that they charge customers for the sugary food product, to PC. This is the new (higher) price that consumers pay for the product. The manufacturers receive that price, but then must pay the tax to the government, and the quantity of the good that is traded falls to Qt (note that this is likely to be the point of the sugar tax - to reduce consumption). The effective price that the manufacturer receives (after deducting the tax) is the lower price PP. The difference between PC and PP is the per-unit value of the tax. However, the important point here is that, even though the manufacturer is the one paying the tax to the government, the consumers pay part of the tax (the difference between PC and P0), and the manufacturers pay part of the tax (the difference between P0 and PP). It isn't possible to mandate that the manufacturers face the burden of the tax alone. In the diagram above, the manufacturers and the consumers have roughly equal shares of the tax burden.

In fact, it might be worse. There are some that argue that sugary products are addictive. If this were the case, we would expect demand to be relatively inelastic (demand would not respond much to price, because addicts will continue to buy roughly the same amount even if prices rise a lot), which leads to a demand curve that is relatively steep, as in the diagram below. Now think about the share of the tax burdens in this case. Notice that the price for consumers (PC) has increased significantly over P0, while the price for manufacturers (PP) has barely changed. The burden of the tax falls mostly on the consumers! And to make matters worse, the quantity of the sugary food product traded barely changes at all (falling from Q0 to Qt).

So while a sugar tax might be effective in reducing consumption, if sugary products are indeed addictive it would take a large sugar tax to have an effect on obesity and health (unless people switch from sugary food products to fatty food products instead!). Note that large excise taxes are what we currently apply to tobacco.

Finally, what happens if the tax is levied on consumers instead of manufacturers? To do this you would have to find some way of making the consumer pay the tax directly (rather than having it collected by the seller). One way is to require consumers to buy a time-limited licence (or some other document) allowing them to purchase a stated number of units of sugary food products. They buy this licence from the government (which is their tax payment), then take the licence to a seller to buy the product. Sellers would only be legally allowed to sell to licence-holders, and only the quantity that the licence permits. In diagrammatic form, it looks like the market below.

As before, with no tax the market operates in equilibrium with the price P0 and the quantity of sugary food products traded is Q0. With the tax, the net benefit of purchasing is lower for the consumer - this is similar to a decrease in demand. We represent this with a new curve (D-tax), which represents the benefits to consumer of the sugary food product (D) minus the cost of the tax (or licence to purchase). In effect the D-tax curve is vertically below the demand (D) curve by exactly the per-unit dollar value of the tax.

Note that the effects of the tax look identical to those in the first diagram in this post. The consumers pay a higher effective price (PC, which is made up of the PP they pay to the manufacturer, plus the amount paid to the government), the manufacturers receive a lower price (PP), and the government pockets the difference in prices. Quantity traded falls from Q0 to Qt.

However, we don't typically see taxes on consumers (rather than producers) because they are more difficult and costly to collect. It is relatively easy and cost-effective to collect taxes from sellers, because there are fewer of them (than buyers), and because they have to advertise where they are (otherwise buyers couldn't find them) it is easy for the tax collectors to find them.

Having said that, if you wanted to really increase the cost of sugary food products to consumers, making them buy a licence to purchase at a local government office before they can go to the store and buy a Snickers bar is certainly going to make many of them reconsider!

What is a developing country, anyway?

I've always thought it problematic that we define "developing countries" (or "less-developed countries" or "low income countries" if you prefer) purely on the basis of income (or GDP per capita). There are many aspects of development that are not captured by income (although they may be correlated with income), such as health, education, good institutions, individual freedoms, and so on. The focus on income is one of the main reasons why we conflate under-development with poverty. While they may be related, they are not the same thing.

The Human Development Index (HDI) goes some way towards improving the categorisation of countries, but isn't really used when the big multilateral agencies consider categorising the countries that are most in need of assistance. Moreover, by its very nature as an index, it is uni-dimensional. What we need is a more multi-dimensional way of categorising the level of development of countries.

So, I was quite excited to recently read this 2014 CGD Working Paper by Andy Sumner (King's College London) and Sergio Tezanoz Vazquez (University of Cantabria). It builds on some earlier work of theirs that was published in the Journal of Development Studies in 2013 (sorry I don't see an ungated version online), and looks at a taxonomy of developing countries constructed in an explicitly multidimensional way.

The authors use cluster analysis to categorise countries, using variables across four main dimensions of development:

1. Development as structural transformation - GDP in non-agricultural sectors (as a % of GDP), exports of primary commodities (as a % of GDP), GDP per worker (in constant 2005 PPP dollars, as a measure of productivity), number of scientific articles (per million people), and external finance (overseas development assistance, foreign direct investment, foreign portfolio investment, and remittances, as a share of GDP);
2. Development as human development - poverty headcount (using a $2 per day poverty line), Gini coefficient (a measure of inequality), and malnutrition prevalence (low weight-for-age among those aged under five years);
3. Development as democratic participation and improved governance - World Governance Indicators index, and POLITY 2 index; and
4. Development as environmental sustainability - CO2 emissions (in metric tons per capita).

Cluster analysis is a really useful way of identifying observations (in this case, countries) that are similar across many dimensions. The important thing is that, unlike the HDI, the multi-dimensionality of the data is preserved. That means that you don't have a dichotomy (poor/non-poor countries), nor do you necessarily have a single development trajectory (from low income to high income). The authors note:

hierarchical cluster analysis allows one to build a taxonomy of countries with heterogeneous levels of development in order to divide them into a number of groups so that: i) each country belongs to one – and only one – group; ii) all countries are classified; iii) countries of the same group are, to some extent, internally ‘homogeneous’; and iv) countries of different groups are noticeably dissimilar. The advantage of this procedure is that it allows one to discern the ‘association structure’ between countries, which – in our analysis – facilitates the identification of the key development characteristics of each cluster.

Moreover, one of the great things about this working paper is that they look at two points in time (1995-2000, and 2005-2010), which allows them to:

...this analysis allows us to... analyse the dynamics of the development process of a single country in comparative terms (that is, in terms of the average development indicators of the "peer" countries belonging to the same cluster.

In both time periods, the authors identify five clusters of developing countries. The three variables in order with the greatest discriminating power (the variables that make countries in each country most different from each other) are poverty, quality of democracy, and productivity in the 1995-2000 data, and poverty, productivity, and quality of democracy in the 2005-2010 data. The consistency is reassuring.

The five clusters in 1995-2000 (in order from lowest average Gross National Income per capita to highest) were:

1. Very poor countries with largely 'traditional' economies - 31 countries, including Democratic Republic of Congo, Rwanda, Pakistan, and Swaziland;
2. Poor countries with democratic regimes but poor governance - 18 countries, including Ethiopia, India, Indonesia, and the Philippines;
3. Countries with democratic regimes but high levels of inequality and dependency on external flows - 18 countries, including Moldova, Honduras, Colombia, and the Dominican Republic;
4. "Emerging economies" that were primary product exporting with low inequality but high environmental pollution and severely constrained political freedoms - 11 countries, including Azerbaijan, China, Egypt, and Gabon; and
5. Highly polluting and unequal emerging economies - 21 countries, including Ukraine, Thailand, Mexico, and Argentina.

In 2005-2010, four of the clusters maintain a similar definition, but Cluster 2 becomes "Countries with high poverty and malnutrition rates that are primary product exporting and have limited political freedoms".

To a large extent, the most interesting aspects of the paper are the dynamics, i.e. which countries move from one cluster to another over the period, and which countries remain in the same cluster. There is a lot of movement between clusters - too much to effectively summarise here. Some notable (to me!) movements though include Vietnam moving from Cluster 1 to the new Cluster 2, India (and Nigeria, Ethiopia, and Papua New Guinea) moving from Cluster 2 to Cluster 1, Indonesia (and Sri Lanka) moving from Cluster 2 to Cluster 3, Thailand (and Ukraine) moving from Cluster 5 to Cluster 3, and Iran moving from Cluster 5 to Cluster 4. All countries in Clusters 3 and 4 in the first period were still in the same clusters in the second period, which is also interesting.

Obviously, some improvements can be made in terms of which variables should be included in developing the clusters (the authors make this point themselves). However, I can see a lot of mileage in further exploring not only the results in this paper, but the approach to categorising developing countries and their development paths more generally.

Where Radiohead went right, and Nine Inch Nails went wrong

In ECON110 every year we have a half a topic on the economics of media industries. Part of that topic is considering new revenue stream models for the media industries that are most 'at risk' from digital downloads (i.e. those industries that are most reliant on revenue from content, as opposed to revenue from selling audiences to advertisers).

One thing I always cover is the Name Your Own Price (NYOP) movement in the recording industry. There are a number of notable examples, like the Canadian singer-songwriter Jane Siberry, who allows downloaders to choose the price they pay for any song they download - if they choose to pay nothing, it is "a gift from Jane" (although I note now that her site is down?). Other examples include Radiohead's almost-free pre-release of their 2007 album In Rainbows, where Radiohead released the album on their website where fans could pick their own price. The pre-release was available from October 10 to December 10, 2007, after which the album was released in digital and physical forms in early January 2008. The pre-release was 'almost-free', because even if someone was to choose to pay zero, they had to pay a service charge of £0.45, plus they had to provide their contact details (a non-monetary cost).

A paper earlier this year (ungated earlier version here) by Marc Bourreau (CREST-LEI), Pinar Dogan (Harvard) and Sounman Hong (Yonsei University) looks specifically at the impact of Radiohead's strategy. Specifically, they look at the possibility of two effects occurring: (1) a 'cannibalisation effect', whereby the early online sales reduced later sales (that would have occurred had the pre-release not happened); and (2) a 'buzz effect', whereby the pre-release generated a lot of media attention, increasing the profile of the album and increasing overall sales.

They compare the release of In Rainbows to the releases of a control group of bands similar to Radiohead, based on a variety of criteria (the control group was Sigur Ros, Muse, Arcade Fire, Beck, Interpol, Bjork, Coldplay, The National, Arctic Monkeys, Franz Ferdinand, and Sonic Youth). They find:

In the case of conventional album releases, the estimated coefficient at the "peak" is about 2 for digital album sales... This is translated as an increase in the digital album sales, which are about 7.4 times higher in the week of the release compared to the previous week...

In the case of the PYOP pre-release strategy of In Rainbows, the coefficient in the "peak" is estimated to be about 3.7 for Radiohead's digital albums. This estimate suggests that Radiohead enjoyed an increase of about 40 times in its digital album sales compared to the week before.

It's easy to see that there is a substantial difference between a 3.7 times increase in sales, and a 40 times increase in sales. On the flip side, there was little impact of the digital pre-release at all on CD sales (they neither increased nor decreased as a result). The authors conclude that the pre-release strategy generated a substantial 'buzz effect' for Radiohead, that more than offset any 'cannibalisation effect' on digital sales. They provide additional evidence for this in the form of Google Trends searches for Radiohead, and mentions of Radiohead in print articles (from Factiva).

What about Nine Inch Nails? They released their 2007 album The Slip for free download from their website (it's still available there now). There wasn't nearly as much hype about this release as for Radiohead (according to Google Trends and Factiva), and Bourreau et al. find that the release had a negative effect on NIN's digital album sales (when compared with their control group: Marilyn Manson, A Perfect Circle, Tool, Queens of the Stone Age, Deftones, KoRn, Smashing Pumpkins, Massive Attack, Rob Zombie, and The Prodigy). In other words, the cannibalisation effect outweighed the buzz effect.

Overall then, a win for Radiohead (in more ways than one - In Rainbows won two Grammy awards). NYOP works, but only if you can generate enough hype to ensure that sales increase overall. Or, it might work also for new artists who are trying to overcome the information asymmetry problem (potential listeners have no idea if they are any good or not) - this is the model that bandcamp.com is founded on.

Who cares about income inequality?

The provocative title to this post is taken from an article in Policy Quarterly (pdf) by Philip Morrison (Victoria University) earlier this year. On the surface, it seems like a straightforward question - most people care about inequality, right? However, once you dig into the data a little bit (as Phil has done), you find that it's actually a very reasonable question to ask, with somewhat surprising results.

In the article, Phil used data from the World Values Survey (1998, 2004, and 2011) and from the International Social Science Programme (1996 and 2006), and looks at the questions on attitudes to income inequality and redistribution.

The World Values Survey:

asks respondents to consider whether ‘Incomes should be made more equal’, a response of 1 denoting complete agreement. At the other end of the scale is the statement, ‘We need large income differences as incentives for individual effort’, with a 10 indicating complete agreement.

I'm not convinced that there is a true dichotomy between those responses - it might be possible to strongly believe that "incomes should be made more equal" and that "we need large income differences as incentives for individual effort", if one believed that the existing large income differences are too large, for example. That problem aside though, the results are interesting. Excluding the "don't know" responses, in 1998 a slightly higher proportion of New Zealanders favoured more equality (47.2% answering 1-4 vs. 46.5% answering 7-10 on this question). However, by 2004 this had switched to fewer favouring more equality (45.9% vs. 48.9%), before switching back in 2011 (50.9% vs. 42.3%). Given that these survey results span the period before and after the financial crisis, Phil suggests that these results support "the view that an increase in economic growth lessens pressure for government redistribution of income".

In the International Social Science Programme, the question asked was:

‘What is your opinion of the following statement: “It is the responsibility of the government to reduce the differences in income between people with high incomes and those with low incomes”.’

It used a five-point Likert scale in 1996, and a four-point Likert scale in 2006. On the results, Phil writes:

A comparison of the responses to the 1996 and 2006 ISSP surveys... suggests a reduction in support for redistribution over the intervening decade, a result which is consistent with the apparent decrease in preference for greater equality observed over the first two WVS surveys. In 1996 only 43.37% disagreed or strongly disagreed that it is government’s responsibility to reduce income differences between the rich and the poor in New Zealand. By 2006 this had risen to just over half, to 50.21%. The fact that both surveys were administered first in high and then in low unemployment periods likely accounts for at least some of the shift in attitudes towards income inequality.

International comparisons show that New Zealand (along with the U.S.) is somewhat of an outlier though:

In both survey years New Zealand respondents were among the most likely to hold the view that it was definitely not their government’s responsibility to reduce income differences, the 20.7% in 2006 being exceeded only by the US at 21.1%... the fact that we are now much less likely than most other countries to support income redistribution is a feature that surprises many older New Zealanders nurtured in the welfare state.

New Zealanders' shifts away from a belief in redistribution are also observed in the U.S. Alongside this shift in the U.S. has been a shift towards fewer people believing that health care is a right. Hopefully, we aren't following that track.

Overall, I find the results quite surprising - only about half of New Zealanders are in favour of lower inequality, and a smaller proportion of us (that those in other countries) are in favour of redistribution. My argument is that fewer people care about inequality than care about poverty. As an illustration of this, I often pose this thought experiment to my students: Would the problem be reduced by burning 10% of the wealth of all of the richest people? If the answer is yes, then the problem probably stems primarily from inequality. Otherwise it is more likely to be primarily a problem of poverty. Most of the time, the problems that people are concerned about are problems of poverty, not inequality (as illustrated in the Max Rashbrooke book "Inequality: A New Zealand Crisis"). Or often we care about global inequality, which is conflated with poverty (as this John Cochrane post illustrates).

So, who does care about domestic inequality? Phil concludes:

New Zealander’s do care about income distribution and the role their government should play. However, what they care about differs markedly. The survey results presented above indicate a fine balance between those who would like to see less and those who would like to see even greater income inequality.

The open question now is "Why?".

The unintended consequences of Wikipedia nudges

Nudges are one of the 'in things' at the moment, for business and policy. We have witnessed the rise of the Behavioural Insights Team (a.k.a. the Nudge Unit) in the U.K. and the Social and Behavioral Research Team in the U.S., and even the Australians (NSW Department of Premier and Cabinet) have a Behavioural Insights Community of Practice. But do nudges always work as intended?

Not according to a post earlier this week on the Misbehaving blog, by Jamie Kimmel. Of the three examples he provides, my favourite is the Wikipedia example:

Wikipedia began testing new ads on its website to bring in more revenue via donations. The central theme of these messages is around two facts: Wikipedia has 450 million users, and less than 1% of those readers donate to the site.

Problems arise, however, when we look at evidence from behavioral science around the use of normative statements. This one from the context of voting: 

"The fact that many citizens fail to vote is often cited to motivate others to vote. Psychological research on descriptive social norms suggests that emphasizing the opposite—that many do vote—would be a more effective message. ...Practically, the results suggest that voter mobilization efforts should emphasize high turnout…. More generally, our findings suggest that the common lamentation by the media and politicians regarding low participation may undermine turnout."

So it’s possible that, by pointing out how few donate, these Wikipedia messages may actually be counterproductive.

We don't have any evidence to support the failure of this nudge, but it does seem plausible that it isn't working as intended (and may actually be counter-productive). Quasi-rational people are affected by framing. If less than 1% of users donate to the site, that might still be millions of donors. Thus, it might be more effective to use a positive framing: "Join the 2 million others who have donated to Wikipedia this year!" or something like that.

Having said that, as you would expect it appears that Wikipedia is using some A/B testing to determine the most effective strategy for generating donations, according to this article:

Settling on the best format for ads is an ongoing challenge for the fundraisers at Wikipedia, who test different layouts throughout the year to see what gets the most engagement. Ads that garner most traction are then prioritized for the December campaign.

So, perhaps the negative framing is working? Or maybe they just haven't tested a positive framing yet. I guess we will find out next month when Wikipedia starts its annual fundraising campaign.

Is love becoming less important over time?

Yesterday I read an interesting paper (sorry, I don't see an ungated version) published back in 2000 in the Journal of Bioeconomics by Alan Collins (University of Portsmouth). In the paper, Collins econometrically analyses the extent to which people lose their virginity in love. Overall the paper is pretty shallow - the discussion of the results and conclusion is very brief, and the sizes of coefficients are not discussed at all (Ziliak and McCloskey would not be impressed). However, I think it does provide at least one interesting result that deserves comment.

Collins uses data from the U.K. National Survey on Sexual Attitudes and Lifestyles from 1990-91. The sample covered adults aged 16-59, with a reasonable sample size of over 4,500 responses. The main results are based on a regression analysis where the dependent variable is whether a respondent gave 'being in love' as the reason for their loss of virginity.

It turns out that there are very few explanatory variables that are associated with losing virginity in love - education, how the respondent learned about sex, belief in God, and religion were mostly insignificant (the exception being that Muslims had 73 percent lower odds of reporting losing their virginity in love than the control religious category (those who were not Catholic, Hindu/Sikh, Muslim, or Jewish).

The interesting results are the differences between men and women, in terms of the effect of age. In the paper, Collins misinterprets his results as showing differences between women when younger and when older. However, because all respondents were surveyed at the same time (1990-91), then age is really a proxy for cohort effects (we can't separately control for age at loss of virginity, and the year that virginity was lost, but assuming that most people lose their virginity about the same age, i.e. not decades apart, then this probably isn't too far from correct). These cohort effects tell us whether love is becoming more, or less, important over time in the virginity loss decision (it doesn't tell us differences between younger and older people at the time they lost their virginity). Because respondents were aged 16-59 at the time of the survey (average age around 35), then the time period (for loss of virginity) covered is approximately the late 1940s to the late 1980s.

In terms of results, it turns out that love has become no more or less important for men in the U.K. over this time period - each additional year of age is associated with a 0.1 percent decrease in odds of reporting that their virginity was lost in love (and the effect is very statistically insignificant). For women though, each additional year of age is associated with a 1.6 percent decrease in odds of reporting that their virginity was lost in love (and statistically significant). So, for women love is becoming less important for virginity loss over time (extrapolating over the 40-year period 1940s-1980s, a 40-year difference in age equates to a 48 percent lower odds of reporting losing virginity in love).

This result has the effect of reducing the difference in responses between men and women (women had 3.4 times higher odds overall of reporting losing their virginity in love than men). Sadly (for the romantics among us), this tells us that love is probably becoming less important over time, for women.

Read more:

• Is sexual freedom a normal good?
• More sex might increase happiness, but not if an economist tells you to
• Are pornography and marriage substitutes for young men?
• Economics majors have the most sex!
• Is sex like driving? HIV prevention and risk compensation
• An un-recognised additional cost of higher education