I shall deal with the points made by Denver and Hands in the order in which they raise them.

(1) Since my article tries to show that the relationship between marginality and both turnout and changes in turnout is, at best, a faint one, the assumption that ‘parties are more efficient in, or put greater effort into, marginal wards’ is in no way necessary to the argument. On the contrary, the assumption seems to be characteristic of those who claim a relationship between marginality and high turnout.

Denver and Hands point to accumulated evidence that concentrated efforts can have a dramatic effect on turnout. It would be foolish to deny that this can be the case or has, indeed, been the case, but the figures seem to suggest that the effect is usually small to insignificant. Most of their British evidence is based upon general not local election results. While there is clear evidence to show that turnout increases when a constituency becomes marginal (Brighton, Kemptown in 1966 is the classic case), the evidence about local elections is muddled and inconclusive, and has often been used to support conclusions which do not follow.