Every year one million married couples in Europe decide that they cannot stand the sight of each other and split up. The Institute for Family Policy has just presented this dismal statistic to the European Parliament and called for action. It is hard to feel much confidence in the outcome: the institution of marriage and the institution of the European Parliament are as hopeless as each other. However, I can think of one thing that might improve the former. That would be to overhaul the market for prospective spouses, bringing it in the 21st century and making it operate a bit more like any other. Bob Reid, ex-head of British Rail, once said that people who got divorced either had a retention problem or a selection problem. Either it was the tedium of listening to your spouse tell the same old anecdotes decade in, decade out, or it was the fact that you had failed to pick the right spouse in the first place. Of the two, I suspect the selection problem is the most lethal; luckily, it is the easier to fix. At the moment we pick our partners in the most slapdash fashion. We rely on sexual attraction and on a basic hunch to tell us if the person is a good egg. The first does not usually last and the second is unreliable. To make matters worse, we are often not in our right minds when we choose. They say that love is blind. Actually it is more like being on drugs: scientists say the chemicals released in the brain in the first flushes of love have a similar effect to cocaine. Do we let people off their heads on cocaine make important decisions? Of course not. So why would we let them screw up their lives by choosing the wrong spouse when in no fit state to decide? The decision process is further hampered by our refusal to do due diligence. When I bought a pair of jeans on Ebay the other day I got detailed information about the trustworthiness of my trading partners ("great e-bayer!" "smooth transaction"), which was far more than I ever asked of my husband when I started dating him. When a company hires anyone they leave no stone unturned: interviews, references, psychometric tests and medicals. Equally, when we decide to buy shares on the stock market we have more information than we know what to do with: how much profit the company has made in previous years, past share price movement and so on. But in matters of the heart there is no information freely available, and we seek out none. Following other transactions we are expected to leave feedback. When we stay in a hotel or get our tyres changed or go to the doctor we are asked to complete forms to rate the service. But after going on a date our experience, which would be of great value to future dates, goes unrecorded except on Facebook, where the wronged leaves the occasional spiteful post. In the old days the business of choosing a mate was less risky as we lived in cosy communities in which everyone knew everything about everyone else. In most of Jane Austen's novels the heroine is poised to marry a cad, but then - just in the nick of time - discovers a damning past. Anne in Persuasion might have married the awful Mr Elliott if her friend, Mrs Smith had not known him and was able to tell her about his grasping, shoddy ways. What the marriage market needs is an online rating agency that would collect detailed feedback on individuals as prospective love partners. Date Rate, as it could be called, would be a cross between Ebay and Wikipedia. Former lovers would provide information covering such relevant factors as fidelity, sexual appetite, generosity, dedication to watching football on television, tendency to leave dirty socks strewn around and so on. And, as on Wikipedia, there would be a certain amount of biographical detail as well as scope to change inaccuracies. If my system had been in operation millions of disastrous marriages would never have happened. Princess Diever would have married Prince Charles, as she would have quickly seen that he happened to be in love with someone else. And Paul McCartney certainly would not have married Heather Mills. The system would have the beauty of being impersonal. One of the most difficult dilemmas a friend or parent can face is whether to tell someone that they are about to make a colossal mistake in their choice of spouse. Date Rate would make such dilemmas vanish, as negative information would be already posted on the site. There are three snags with Date Rate. First, it is not romantic. But then romance has not proved a particularly good model so far. Second, it is an invasion of privacy. But then one could perhaps get round that by giving each person the power to refuse access to their own information. The worst problem is that it penalises the reformed love rat, whose Date Rate card looks so bad that no nice suitor will have him. But, then, as most love rats do not reform, maybe that is a price worth paying. In spite of these drawbacks, Date Rate has a redeeming feature that is common to all the most efficient, transparent markets. It would encourage people to behave better. If you knew that treating your partner in a beastly fashion would be taken down in evidence and held against you by future partners, you might even think about behaving more decently in the first place. By Lucy Kellaway, Financial Times lucy.kellaway@ft.com