Gambling as socially useful: weight loss betting

A pound of flesh?Alongside prostitution, gambling is one of those moral issues almost guaranteed to evoke visceral responses. It cuts to the core of the eternal socialist/libertarian debate – do we allow, discourage, or outlaw an activity with such obvious negative consequences for individuals and families?

In some cases, however, gambling might not be as bad as it’s cracked up to be. I’ve written before about using betting markets as prediction tools, something that I think will continue to grow in the future. The success of such attempts hinges on the power of monetary gain as a mechanism for yielding individuals’ best efforts. Recently I came across another initiative using the same underlying incentive idea, but with a different purpose: betting on your own weight loss.

Here is one diet casino site, although the simplicity of the idea makes me think that there are probably at least a handful floating around the web. The arrangement is quite simple: set your weight loss target, choose your bet, and provide some self-info. Then the company will tell you how much you stand to win, and check up on your progress at regular intervals (with a mechanism to prevent cheating).

The theory behind this is straight out of behavioural economics – betting is an excellent commitment device. Humans use these devices all the time to bridge the gap between long-term payoffs and short-term pain. Dieting, studying for tests, or being nicer to idiot co-workers all yield dividends sometime in the nebulous future, whereas the associated suffering happens now. The temptation, therefore, is to allow ourselves to cheat. Commitment agreements impose some sort of penalty for doing so (e.g. forfeit your stake in the weight loss betting scenario), to counter the short-term temptation. There is a good paper reviewing commitment devices here.

If it helps people lose weight, it must be a good thing, right? Well, weight loss is certainly a noble goal, and socially useful (fill in your favourite alarming statistic about obesity here). There’s more to it than that, however – some of my thoughts on the idea:

  • Returning to the casino analogy, the odds must be stacked in favour of the company. Although the site I mentioned early claim that they make money from “corporate and government clients who are interested in creative solutions to weight loss”, they still want to make a profit (on average) from each client. Thus those questions you are asked at the beginning (which include the strategy you are planning to use, your reason for wanting to lose weight, your gender and your height) will give the company’s actuaries a pretty good idea of how likely you are to succeed, and therefore how large the prize you are offered should be. Apparently in a normal casino the house edge ranges from 0.5% to 20%+, but since people have the dual goals of fiscal gain and weight loss in this scenario, I would be surprised if this company doesn’t skim even more off the actuarially fair payoff.
  • This flies in the face of standard microeconomic theory regarding risk aversion. Weight loss is an uncertain outcome, so you may end up disappointed or happy. Far from wanting to compound this by then betting on their desired outcome, risk averse consumers should in fact want to take out insurance against the possibility of them not achieving their goal. Any such insurance market would suffer from huge moral hazard problems, though, which is probably why (to the best of my knowledge) no such one exists…
  • Anecdotal evidence suggest that programs with one big weight loss goal and associated payoff (e.g. ‘The Biggest Loser’ TV show) work quite well to induce people to hit that target, but that the improvement is difficult to sustain – i.e. individuals often return to their previous weight. My guess is a crowding-out effect – the extrinsic motivation reduces the individual’s need to develop the intrinsic motivation that would have helped them continue the good habits they began while working towards their goal. I’m guessing that these weight loss bets will suffer from the same problem.
  • Further to the first point above, there’s a bit of an incentive misalignment operating. Individuals are signing up with a company that promises to help them with their weight loss goal (even going to far as promoting the services of their ‘Chief Fun Officer’ in doing so). The company only makes money, however, when their clients fail – something of a problem there.

If anyone has tried such a program as this or knows someone who has, I’d be fascinated to hear about your experience – tell us about it in the comments section below.

What I’m browsing 18/12/13

Some links for your internet pleasure:

Monopoly for finance enthusiasts

Did your family go down the Monopoly road? I think you can tell a lot about family dynamics from measuring Monopoly frequency, at least when the children are between the magic ages (between 8 and 15, by my reckoning). The game is notorious two things: causing fights and lasting for hours. I recently read an article suggesting how to make the game more interesting – go check it out.

My thoughts on Finance Monopoly:

  • Only the British Monopoly - don't give me this crap about 'Boardwalk'!A few more rules are probably needed to clarify the suggested additions. For example, order of debt repayments (does rent owed take priority over debts from previous turns? who gets first claim on a bankrupt player’s assets?), and timing of financial transactions (can you only deal on your turn, or does the engine of finance march on relentlessly?) are two that spring instantly to mind.
  • Boy would these rule changes allow the game to get fiendishly complicated, fast. Not sure about your typical family game on a Sunday afternoon, but if we played this at an Economics and Finance Society meeting, the game would probably last at least a few days.
  • On that note, there could be a role for a dedicated banker and ‘government’ – to hand out money, administer contracts, etc. I actually dreamed up a Monopoly variant like this a few years back, where players would be able to take on roles like ‘builder’ (gets the money from any construction) and ‘tax collector’ (gets a cut of any taxes). Watch this space – I may write a post about it at some stage…
  • For some added game theory fun, the rules could explicitly state that no contracts are enforceable – that is, players are completely free to tell their creditors to go to hell. You’d quickly discover whether a loss of reputation and access to the in-game credit market acted as sufficient deterrents.

My interest is now piqued – I’d quite like to see how a game of Monopoly Finance-style plays out! Anyone got six hours to spare?

Concession tickets on Sydney trains

Cityrail trainSydney Trains (formerly Cityrail) has recently implemented a new policy whereby concession tickets of all stripes – pensioners,students, veterans, the works – cannot be purchased from ticket machines in during certain times. The organisation’s website is cheerfully vague about when these times are, but from my experiences it seems like 9:30am is the cutoff for buying concession tickets from machines.

Now, I love the Sydney Trains ticket machines. They are quick, fairly reliable, and easy to use, plus they are conveniently located at myriad places in busy stations. Buying tickets from a ticket window, on the other hand, is a nightmare. The queues to occasionally stretch out the door of the station, communication is often difficult with the employee (partially due to poor microphone quality), and there are far fewer ticket windows than ticket machines. A combination of the factors I’ve just mentioned has caused this new concession ticket policy to make me miss a train (yes, I tend to cut it a little fine – see the quote in this Freakonomics post for a defence).

The point of the initiative is to cut down on the number of commuters illicitly travelling on a concession fare – i.e. those who should be paying full price. I have a couple of comments about this:

  • I’d love to see the data on how big this problem actually is – my gut feeling is there can’t be all that many people cheating the system, if only because doing so would threaten their self-image as an honest, upright citizen.
  • Failing to produce a concession card when travelling on a concession ticket already carries the same fine as travelling without a ticket, so the people who are doing so will continue only if they are fairly confident of not having their ticket checked by a guard. Increasing the number of ticket checks would help to diminish the problem and carry the spillover benefit of decreasing other offences such as graffiti, alcohol consumption, and anti-social behaviour.
  • The benefits of this policy must be pretty slim, particularly when weighed against the costs to consumers in terms of extra inconvenience. There’s no way Sydney Trains could completely stop people buying concession tickets from machines, as the resulting gridlock at the ticket window would cause chaos – plus it would necessitate a bunch of extra staff being rostered just to meet the demand for ticket sales during the morning peak period. Yet this is exactly the period when the highest volume of travel is done – therefore most commuters are just as capable of buying concession tickets to which they aren’t entitled as before the policy.

In short, this is a silly policy that should be scrapped and replaced with something better, like greater presence of officials on trains to check tickets and quell anti-social behaviour.

Steam sales and discount rates

Bought 'Civ  V: Brave New World' on Steam recentlyI usually labour under the happy illusion that I am among the subset of the population with low discount rates – a trait that is linked to education. A recent experience belied this belief, however – my purchase patterns in the current sales on Steam.

I’m a big fan of the Civilization series of games – I spent many a happy Saturday morning playing Civ II, content in the knowledge that it would be hours before my parents awoke to kick me off the computer. The current iteration of the game, Civ V, is an amazing piece of work – grand strategy at its finest. I intend to write a post on it and other games in the near future, so watch this space.

Anyway, the second expansion pack for Civ V came out a few months ago, and as much as I wanted to jump on board straight away, it was fairly expensive and also I was still writing my thesis. So I waited.

Thesis done, the Steam autumn sale started last week, and lo and behold – the game was on sale! 50% off, what a bargain! One downside, though – popular games like that usually get marked down pretty heavily, so there was the possibility that it would become even cheaper as the sale period continued. After a few moments’ reflection, I decided to get on board and take the half-price offer. That was Friday.

True to Murphy’s law, though, this morning I discovered the game’s price had dropped further to 66% off. The disinclination to wait a few days cost me 16% in extra savings 🙁

So, exam question for you all: given those figures, what is my implied annualised discount rate? For bonus marks, consider imperfect foresight over sale price trajectories.

Stata and parentheses

Just ordered a personal copy of Stata 13 (I wanted to go all out and get Stata MP, but the little voice of reason in my head convinced me to buy a one-year licence for Stata IC instead), and two things have jumped out at me already.

  1. Man, it runs nicely on my new iMac. The computers at UOW weren’t too bad, and the IT department there keeps them on a reasonably generous upgrade cycle, but top-notch new tech makes a hell of a difference.
  2. This might be a new feature in Stata 13, or it might be due to the OS change (I used Stata in a windows environment for my honours), or I may just not have had the good sense to find and activate the feature before, but in my .do files selecting a parenthesis causes its partner to be highlighted in red. Seems simple, but already in my short life as an economist I have spent a non-trivial amount of time debugging my code only to discover that I either omitted or inverted a parenthesis somewhere. This small feature should therefore make a difference to my mental health in the long run.

Any other nice aspects of Stata 13 or Stata on a Mac OS that I should know about? Let me know in the comments section.

Vocabulary and signalling

Before we start, go and check out this link: Test Your Vocabulary.Daniel's known vocabulary (from the site)

As an economist, anything with a sample size in the millions (!) has me drooling, even if a test administered over the internet is going to be unavoidably biased. The idea behind what they’re doing – constructing a measure to predict someone’s vocabulary from observable characteristics – will, if nothing else, highlight any flaws in tests like the SATs in the United States.

Tests such as the SAT and GRE aim to measure how likely an individual is to succeed at university study. Evidently, success is related to factors such as discipline, intelligence, existing knowledge, etc., all of which are difficult to measure. Instead, the tests shoot for something easier: vocabulary size. The underlying assumption is that individuals with larger vocabularies will be smarter, harder working, more knowledgeable – in short, possess a larger dose of the unobservable qualities that make for a good university student.

Trouble is, it’s widely known that these tests feature a vocabulary component. Individuals who therefore are not the best cut out for university can attempt to appear as though they are by developing their vocabularies specifically to ace these tests. To spell it out more clearly: universities want a certain calibre of student, which they attempt to identify. Individuals who are not in this group can still falsely signal that they are by improving their abilities to score high marks in the test (rather than actually improving their general abilities and hence moving into the group of desired students).

The data from the ‘test your vocab’ site might help to shed some light on this – it’s a lot harder to fake your abilities in the real world, so the correlation between vocabulary and earnings (a much better proxy for ability, but one which is typically observed after an individual has concluded their studies) will show whether knowing lots of words is a good signal for general intelligence.

Secondly, it will also be interesting to discover some of the good predictors of vocabulary size. Using vocabulary as a proxy for ability isn’t much good if some demographic characteristic (e.g. ethnicity or socioeconomic status) is a stronger determinant of it than underlying capability for success. For the record, that’s my bet. I’m not willing to throw vocabulary out entirely as a proxy for ability, but I predict that it is actually a much less accurate measure than universities hope.

Airline Price Discrimination

Whenever I explain price discrimination to one of my microeconomics classes, I have two go-to examples: cinemas (think students and pensioners) and airlines. The latter example is a lot richer and more fun to use, since there are so many layers of differential pricing for the same product.

I recently found a post that attempts to unpack some of the elements in the price calculation. A good site to stumble across, both as a consumer of air travel and also as an economist. It’s got me thinking about the different categories into which travellers can be sorted, and therefore charged different prices.

What does that have to do with airlines? Oh, wait...Categories of air passengers

  • Hyper-organised OCD-type – buys tickets as soon as they are released. Will pay a premium to reduce risks and cover all bases. Identified by length of time between purchase date and flight date.
  • Business traveller – my default example of price discrimination. Often need to fly somewhere with little notice, and will hence buy tickets close to the departure date. Won’t be covering the costs out of their own pocket: far less price sensitive than a leisure traveller.
  • Families – probably fairly price sensitive, but will also want to keep the kids and adults seated together. Can therefore be induced to pay a premium for joined blocks of seats. Easy to identify – 3+ tickets in one purchase, mixture of adults and children.
  • Technologically-incapable individuals – may purchase tickets over the phone (!), or could possibly scrape together the technical aptitude necessary to find an airline’s website. Unlikely to be cluey enough to do any extensive shopping around or use a cheap flight search site, therefore will bear a reasonably high price, particularly if it is presented as though it’s a bargain. Again, pretty easy to identify – look for the consumers using Internet Explorer.
  • Affluent flyers – if your wage is high, the opportunity cost of your time is high, and therefore any extra time spent travelling will involve a big loss of surplus. Airlines cash in on this by whacking a decent surcharge on non-stop flights.

Far from a complete list, but there’s five to get the ball rolling. Suggestions for more types of flyers in the comments!

Social science research = big(ger) data?

Google Trends - ResearchPaul Frijters is an economist from UQ who likes to ponder the mysteries of life, and one of his blog posts concerns the great existential question facing social scientists everywhere – why are we here? More specifically, what is the point of all of the social science research being done?

Note that we’re not talking about the debate over whether social sciences deserve the label ‘science’, or if obtaining a PhD is worth all the effort. Frijters’s soul-searching query cuts to the heart of life as a researcher: does the knowledge I am producing actually make any difference to the world?

The question arises from an awareness of the vast quantities of research done every single day. With so many results constantly being tabulated, published, and disseminated, how does any one article stand a chance of even attracting the attention of others in the discipline, let alone having an effect on policy? Frijters ponders aloud whether social scientists may, in truth, be writing for the benefit of some imaginary deity with the capacity to comprehend and use all of the vast quantities of insights. More depressingly, he postulates that perhaps we social scientists are simply earning a crust and ignoring the vast elephant-shaped emptiness in the room.

Happily, I disagree with the proposed answers to his question. Instead, I suggest an alternative: research findings are like single observations in an overarching dataset, and this will be yet another area in which man + machine will make great gains over unaided humans.

That is, instead of thinking of the results from research as being the end product, useless unless converted into further research or policy, they should be considered as pieces of a puzzle – perhaps intrinsically interesting, but far more so in conjunction with all of the other pieces. And this is where the technology + human team comes into its own. Machines can quickly sort through and find patterns in the super-dataset comprised of all of the individual findings. Humans are useful at either end of this process – suggesting areas for the machine to concentrate its search, and making sense of what it turns up.

So don’t despair, social scientists – your hard work is contributing to a better tomorrow, just in a more elementary way than you may have expected…

Kitchen equipment, opportunity-cost style!

Pseudo-confession time: I love cooking. And when I say cooking, I mean roasting, baking, sautéing, frying – you name it. As a lexicographically well-endowed person, it is my eternal shame that I use the word ‘cooking’ to refer to all of the aforementioned techniques, but for this I blame my family.

Cookies in Daniel's kitchen (still baking)

The struggle for me over the last few years has been balancing my culinary inclinations with my limited budget as a student. As much as I’d love to have the latest all-in-one super kitchen gadget, there were more pressing budget items, such as the food itself. Opportunity cost strikes again! So with my (extremely) finite resources, I had to prioritise which items were most important to my dream kitchen, and which would have to go on the backburner.

For the benefit of all of my fellow suffering student-chefs out there, here is my top five list of kitchen implements that are worth the investment. To clarify, I’m assuming you already have the mundane items such as cutlery, plates, etc.

Daniel’s Top Five Kitchen Items

  1. Good chef’s knife. The capacity to dismember things is vital in any kitchen, and precision is required. Decent knives seem a little expensive at first glance (a single one can set you back anywhere between $20 and $100), but remember that price is often signal of quality – that holds true here. Plus, investing in a decent knife will not only pay dividends in terms of food quality, it will also save you time and mental anguish. Take it from someone who suffered through blunt, crap knives for years.
  2. Quality frypan. Pans can be obtained pretty cheaply, and for most of your needs the seven-piece set from Kmart will serve you just fine. For messy, sticky jobs like dahl, frittatas, or eggs, you want something non-stick with a nice, thick base. Again, partially this is for quality reasons, but it is more for your health – both physical (cheap frying pans tend to impart bits of themselves onto your food) and mental (washing up a teflon pan is like bathing an angel compared to scrubbing for seemingly hours at its dreadful predecessor).
  3. Slow cooker (aka Crockpot). Boy, do I love my slow cooker. It does everything from soups to casseroles to slabs of meat, and has two big advantages: volume and intertemporal substitution of effort. The volume part speaks for itself (my slow cooker holds 5.5 litres, which usually does four days of dinners for two), and by intertemporal substitution of effort I mean you can do all of the hard work at the start of the day while you’re still fresh, and then come home to a cooked dinner and delicious-smelling house after uni or work. Plus they’re pretty cheap these days, so the cost/benefit ratio is comfortably in its favour.
  4. Electric mixer. This one is more for the baking-enthusiasts out there, since it doesn’t add much to main-meal preparation. As far as the production function for things like pancakes and choc-chip biscuits goes, though, substituting this small piece of capital can save you a whole lot of labour. In fact, I’d go so far as to say that if you want to delve into fancy creations like pavlovas or crèmes brûlées, an electric mixer is a sine qua non.

Actually, it turns out that I can only think of four indispensable kitchen items – with the above, plus the usual gear such as bowls, spoons, oven (don’t laugh – one of the apartments we inspected recently didn’t have one!), etc., I’ve made it through my student years in pretty good culinary and dietary shape!

What are your opportunity cost kitchen implements? Share your thoughts in the comment section 🙂