- Incentives are the cornerstone of modern life.
- Knowing what to measure and how to measure it can make a complicated world less so.
- The conventional wisdom is often wrong.
- Correlation does not equal causality.
- Data shows that past the first generation, a family business does better to bring in an outside manager.
- While free, unlimited, lifetime healthcare is praiseworthy, the economics are tricky. When people do nto pay the true cost of something, they tend to consume it inefficiently. The worried well crowd out the truly sick, waiting times increase and a massive share of the costs go to the final months of elderly patients' lives, often without real advantage. (They did not come up with the answer to the conundrum, though.)
Violent crime in the US suddenly climbed in 1960; the homicide rate doubled but in the early 1990s it began to fall and kept falling. Why? Data shows tough new gun laws had no effect; there is a weak relationship between economic cycles and violent crime. While it helped that there were more police, more people in prison and a decline in the crack heroin market, Levitt and Dubner identified the legalization of abortion as a key factor - fewer children grew up in difficult circumstances. [My note: There is also evidence that violence and violent crime has decreased since the banning of lead in petrol.]
Heart disease is far more common in black people. A possible explanation is that slave traders looked for slaves who could more easily withstand the risks of dehydration (people who could retain more salt and therefore more water). By selecting for one trait, there can be unintended consequences on other aspects of health - in this case hypertension and cardiovascular disease.
Americans are bad at saving money but spend a lot on lottery tickets which typically pay put only about 60% of the take. A prize-linked savings account (PLS) allows you to deposit $100 in a bank account with a deposit rate of say 1%. A quarter of that interest is pooled with the same amount from fellow PLS depositors, and then periodically paid to a randomly chosen winner. Even if you never win the PLS lottery, you keep your original deposit and your 0.75% interest. While some states are experimenting with such accounts, most US state law allows only the state to run a lottery and federal law prohibits banks from operating lotteries.
Incentives can backfire, so find out what matters to people. Types of incentive: financial, moral, social and 'herd mentality'. An experiment in getting people to use less electricity showed that a notice hung on doorknobs saying 'join your neighbours' (herd mentality, which also influences what we buy, where we eat and how we vote) was the most effective. Moral incentives don't work as well as might be imagined. Notices warning people not to steal petrified wood from forest trails in the US actually increased the thefts; people thought 'there's not much around so I'd better get my bit now'.
Charity fund-raising - getting new donors is difficult and expensive. Donating makes people feel good, but they also feel bullied by being asked. A charity used 'once-and-done' mailing acknowledging how people might feel and asked donors to tick one of (a) this is my only gift, please do not contact me again, (b) please contact me only twice a year or (c) please keep me updated with regular communications. Result, only a third opted out of future mailings, and the others raised overall donations by 46%.
Helping business. Zappos clothing and shoe store made its customer service 24/7, and posted its phone number prominently on the website. Call-centre work is often poorly paid, so Zappos offered its customer reps fun and power. Company meetings might be held in a bar, work cubicles can be decorated, there is no dress code, reps are encouraged to talk to customers for as long as they want and there is no script, they can settle problems without reference to a supervisor, and can 'fire' a customerwho makes trouble. Result - the company never lacks applicants for call centre work.
Some countries have tried easing traffic congestion by restricting the days each person can use their car. Result: some people buy a second, older car, streets remain congested, no increase in use of public transport, and more pollution from the older cars. A UN incentive plan financially rewarded manufacturers who destroyed stocks of waste greenhouse gas HFC-23; factories world-wide produced more gas to get the credits. Changing the rules meant that the extra HFC-23 produced will likely be released into the atmosphere.
Incentives. (1) What do people really care about (not what they say). (2) Use incentives that are cheap to you but valuable to them. (3) Note if they respond not as you meant - learn and try again. (4) Make incentives co-operative not adversarial. (5) People respond because it is the 'right' thing to do. (6) Some people will always find ways to manipulate the system to their advantage.
A person who is lying or cheating will often respond differently to an honest person. Use their behaviour to identify them. Employers make job application processes easier but often get lots of interest but few with real interest. Colleges and universities make the process hard, weeding out the less committed in the process. Zappos offer new employees after training the option of quitting with a bonus of 1 month's wage, if they have an exit interview and commit to never working for Zappos again. Fewer than 1% of new hires currently accept this offer. Nigerian scammers make their emails so unbelievable that they weed out all but the most gullible.
Persuading people of anything is difficult if they already have a firmly held belief on something. So how can you change behaviour? (1) It's not me; it's you. Your view may be scientifically proved but does not resonate with them. (2) Don't pretend your argument is perfect. Ignore or paper over shortcomings and they will doubt all of it. (3) Acknowledge strengths in their argument. Most people don't take criticism very well and there may be something in their views that strengthens yours. (4) Tell stories (not anecdotes). Stories that are not true can be persuasive, eg. the MMR vaccine caused autism (co-incidence that symptoms start showing around the time of the vaccine administration) or the EU banned overly curved bananas (dreamt up by Boris Johnson.) A US government official needed to create a document setting out what employees were and were not allowed to do. He created The encyclopedia of ethical failure, as a catalogue of screw-ups in the past, with chapters on abuse of position, bribery, conflicts of interest and political activity violation. The stories lodge the rule in your mind.
There is an upside to quitting. People refuse to quit because of (b) being thought a failure, (b) sunk costs (time, money or energy that cannot be recovered) and (c) pay too little attention to opportunity costs. Quitting can be positive. In science, failed experiments identify areas to avoid. Succesful innovation needs to identify the ideas that have potential and those that don't (including good ideas that are simply too expensive at present). Anonymous premortem asks people (e.g. employees, experts) whether a project will be completed in a set time, be a safety hazard, and so on. Their views are likely to be more accurate than those of team leaders or project managers who have a vested interest (financial or prestige) in delivering the project successfully.
Think like a freak by Steven D Levitt and Stephen J. Dubner, Allen Lane, 2014.
