Three years ago, the University of Vermont in Burlington began to experiment with a few nudges towards a healthy, sustainable lifestyle. First, in 2012, campus outlets and the company operating the vending machines were required to make sure that at least 30 per cent of the drinks on offer were wholesome stuff such as vegetable juice, low-fat milk and water. A few months later, selling bottled water on campus was banned outright. The aim, pushed hard by student campaigners, was to encourage students to fill reusable bottles with tap water instead.

So, how did the Vermont experiment go? A study by Elizabeth Berman and Rachel Johnson (of the University’s own Department of Nutrition and Food Sciences) was recently published in the American Journal of Public Health. The researchers found that “per capita shipments of bottles, calories, sugars and added sugars increased significantly when bottled water was removed...As bottled water sales dropped to zero, sales of sugar-free beverages and sugar-sweetened beverages increased.”

In other words, the policy backfired with both barrels. Students didn’t switch to tap water, they switched to the likes of Coke and Diet Coke instead. All this would be just an amusing curiosity — one more example of student campaigners who are all heart and no brains — if it weren’t for the fact that more mature policy makers often commit similar blunders on much broader canvases. We would do well to learn some lessons from the University of Vermont’s experience.

The first lesson is that when it comes to saving the planet, people focus on what they can see. Type “environmental impact of concrete” into a search engine and you are likely to see a page filled with scholarly analysis pointing out that the impact is very large indeed, because cement production releases vast volumes of carbon dioxide. Type “environmental impact of bottled water” instead and your search results will be packed with campaigning groups seeking to persuade you to change your ways.

This is understandable: I can’t do much about concrete but I can stop drinking bottled water. But being a logical target for campaigners is not the same as being a logical target for policy action.

The second lesson is that we often struggle to deal with multiple goals. The University of Vermont wanted to reduce the flow of plastic water bottles to landfill but also wanted to encourage students to be healthy. There’s a clear conflict between these goals. Water is as healthy a drink as you can find, yet that was exactly what the University of Vermont was banning from vending machines. Wishful thinking provides a resolution — if everyone just drank tap water then there would be no problem. But wishful thinking is not an excuse for setting no priorities.

We see this sharply in the debate over nuclear power. We want to reduce the greenhouse gas emissions that result from burning fossil fuels. We also want to avoid radioactive waste and the risk of radiation leaks. In response to a genuine policy dilemma, politicians have tended to plump for wishful thinking every time, typically involving wind turbines.

The third lesson is that the much-vaunted notion of “nudging” doesn’t always help navigate a complicated policy maze. Nudging means using default options, information design and similar techniques to achieve policy goals. It can be very successful. But careless nudges are no more welcome in public policy than at a domino-toppling event. If you pick a questionable target (bottled water) and fudge a key policy dilemma (the environment vs health) then nudging isn’t going to solve your problems.

So what can be done? One approach is to try to reach policy goals with the help of market signals. The classic example of this is a carbon tax, levied on fossil fuels to reflect their carbon-dioxide emissions. The advantage of this approach is that it encourages everybody at any stage of production or consumption to take actions that reduce emissions, because those actions will save them money. A truck manufacturer might develop a cleaner engine, a logistics company might find a more efficient delivery algorithm, and the final consumer might decide to consume a little less.

The idea of using the price system to solve environmental problems is widely accepted by economists but, alas, it finds itself stranded in the policy doldrums. Ponder this: the Pope recently argued that climate change was a grave problem but he opposed market-based responses. Meanwhile the US Republican party likes market-based responses but isn’t so convinced about climate change.

One other advantage of using environmental taxes is that people can decide on their own priorities. A lot of what we do has consequences for the planet — including breathing — and so part of the problem we face is deciding what is worth doing anyway.

Perhaps it is time for a confession. I am writing this column on the hottest July day recorded in British history. At my left hand is a glass of chilled sparkling water, and next to the glass is a plastic bottle to top it up. If there had been a tax on that bottle, it is a tax I would willingly have paid.