I've taken out a contract on myself. Really. It's not that I'm fatally tired of the sound of my own voice (though that time is surely coming to us both). I'm trying to trick myself out of my dismal failure to stick with good intentions.

At stickK.com, I've signed a contract to exercise for four hours a week for 12 weeks. If I fail, the website will donate some of my money to the “pro-life” lobby. So far, my dread of funding bigots to yell obscenities outside abortion clinics has proved a far better motivator than a vague desire for muscle tone.

I found out about stickK.com by reading Nudge, a forthcoming book by the Chicago academics Richard Thaler and Cass Sunstein. It examines why so many of us fail so often to act in our own best interests. Its message is that this is partly because we give so little thought to many of the choices we make. And that education programmes, grandiose incentives or bans may work less well than a cheap and cheerful nudge that simply changes the way choices are presented.

Take airport toilets, which apparently have a big problem with men peeing on the floor. A conventional solution might be posters, or an expensive redesign. But just painting a black fly inside the urinal, for men to aim at, has reduced spillage at Schipol airport, Amsterdam, by 80 per cent.

More serious examples concern financial decisions. It doesn't seem to matter how much information and guidance we are given, very few of us save enough, change the allocations in our pension plans or update our life insurance policies (half of all married people still list their mother as the recipient of death benefits). We tend to opt for the default setting - which in the case of many pension plans means a bleak future.

The solution, argue Thaler and Sunstein, is to design more intelligent defaults. Thaler's “Save More Tomorrow” plan has already boosted retirement savings in America by asking employees to tick a box to increase their contributions every time they get a pay rise. Losing disposable income apparently makes us roughly twice as miserable as a gain of identical size - so this programme ensures that its participants can save without ever seeing a loss of take-home pay. Changing the defaults in this way, working with the grain of human behaviour, has had a bigger impact on savings behaviour than education or information. A manufacturing company which uses the plan found that, after four annual pay rises, joiners had quadrupled their rate of saving.

The way that decisions are framed has enormous impact on behaviour. When fruit is placed in front of desserts in the canteen, children eat better. The order of names on a ballot paper affects how we vote (aspiring politicians should hyphenate with Aardvark). There is no “neutral” design: the designers, often governments, shape lives whether intentionally or not.

So Nudge is an important addition to the “nanny state/no state” debate. We all know about the failures of state intervention. The number of pupils eating school meals has plummeted, we learnt this week, since ministers backed Jamie Oliver's campaign. The Irish decision to tax plastic shopping bags has simply boosted sales of binbags (since many people, like me, re-used their shopping bags for rubbish). Nudge gives more insight into why some policies are self-defeating. It also shows that “no state” is not the alternative as often as we think.

The authors believe that people should be as free as possible to make choices, but that policy should coax people away from the worst. They call this “libertarian paternalism”, a deliberate oxymoron, which the book tries to resolve. Transparency is fundamental to this effort. Citing one study which showed that people were willing to pay twice as much to go to a basketball game if they could pay by credit card not cash, Thaler and Sunstein propose that credit card companies should send annual statements showing exactly what the true costs of the card are. They want people to understand their own weaknesses better.

The book is not just a useful insight into the lazy shortcuts that I now realise plague much of my own thinking. It is also a political manifesto. Both authors have strong connections to Barack Obama: it will therefore raise the hopes of those who would like to think of Senator Obama as less of a statist than he is painted. We know that Mr Obama disliked the levels of coercion involved in Hillary Clinton's plan to force people to buy health insurance. He described his preference for giving people cheaper insurance options as a key philosophical difference between the two campaigns. We also know that he is interested in transparency: he authored a Bill, with a Republican senator, to create a publicly searchable database of federal spending. It would be wrong to overstate his embrace of libertarian paternalism, but it gives a new filter through which to watch him.

I can't promise that Nudge will make you a better person. If you are out to manipulate people, it may be positively corrupting. And that's the worry, really. What is to stop politicians going from nudge to shove? I support making organ donations an “opt-out”, rather than an “opt-in”, a classic nudge supported in this book. I now realise that my position could be construed as an abuse of the simply enormous power of inertia. That said, there is a powerful case for governments to think more intelligently about how they intervene, and about promoting freedom of choice. That I have signed my money over to some silly website is presumably another example of predictable irrationality. But if it keeps me fit, it will have been more effective than any bossy council leaflet. I'll let you know in 12 weeks.

Times Recommends

• Paralysed while the Hamas hydra lives
• Wanted: a superhero to save America
• Blind faith in experts fails justice