Editor's Choice

In praise of trade offs

BMJ 2005; 330 doi: https://doi.org/10.1136/bmj.330.7498.0-g (Published 28 April 2005) Cite this as: BMJ 2005;330:0-g
  1. Jane Smith, deputy editor (jsmith{at}bmj.com)

    Next week Britain has a general election, and over the past three weeks Nick Timmins has been dissecting the manifestos of the three main UK parties (16 April, p 866; 23 April, p 925; p 981). Although sceptical about manifestos—“a newly landed Martian reading Labour's 1997 manifesto and looking at the NHS today would be entitled to wonder how on earth we got from there to here”—Timmins works hard to find real differences between the parties. So do our three commentators (p 986). Indeed, Keiran Walshe argues that the electorate doesn't have much choice: “If you believe in a nationalised NHS or if you want to see health care privatised, you don't have anyone to vote for.”

    Lack of anyone to vote for is a problem because it implies a muted discussion of values and trade offs. The main criticism of this current election is that the parties are claiming the same ground, with no real political arguments about what matters. Inequalities in health and wealth could, for example, be a divisive political issue. The analysis by Mary Shaw and colleagues (p 1016) shows that differences in life expectancy between the richest and poorest groups in Britain have continued to grow. But they have been growing over 20 years, through both Conservative and Labour administrations. Income inequalities have shown a similar trend. The authors make an unashamed plea for redistributive policies and lament that “ ‘redistribution’ is a dirty word in British politics.”

    Yet if politicians fail to come clean about the trade offs, we can always turn to statisticians for some clarity. Robert Hooke's analogy in How to Tell Lies from Statistics between political positions and type 1 and type 2 errors is, of course, meant to help people understand the statistical concepts, but it also helps explain political trade offs. In offering welfare payments, says Hooke, a type 1 error is to give payments to someone who does not deserve them, and a type 2 error is to fail to give them to someone who really needs them. Conservatives tend to find type 1 errors intolerable, and liberals tend to find type 2 errors intolerable. “Only the dreamers really believe that we can eliminate one type of error entirely without allowing the other to increase unreasonably.”

    Or we could look to lawyers. The House of Lords has just ruled that parents who are wrongly accused of harming their children cannot sue doctors or social workers who made or investigated the allegations, even if they behave negligently (p 988). The reasoning is that the professionals' duty is to the child, and in cases of child abuse the child's interest is diametrically opposed to that of the parents. In other words, society finds it more intolerable that a child should be abused than that a parent should be wrongly accused. Clinicians working in child protection will welcome this ruling, but everyone should admire it for not fudging a hard choice.


    View Abstract