Readers who have endured this far will not be surprised to find that our conclusions are largely negative. There is very little evidence that relative shares are constant in the short or long run, but there is some support for the notion that labour’s share has an increasing secular trend. None of the theories of relative shares performs very impressively. Neo-classical accounts of income distribution can be faulted on many grounds, of which reliance on the assumption of perfect competition is probably the most damaging. ‘Degree of monopoly’ explanations are inadequately specified, yielding few if any unambiguous predictions. In so far as neo-Keynesian theory can be formulated precisely, it turns out to be wrong. Finally, ignorance about the effects of trade unions on relative shares remains very substantial.