One thing that has peeved me recently is the claim that falling inflation is good for real wages. This is partially true, and partially false.

It's true in the sense that lower oil prices raise the real incomes of oil consumers. It's also true that a surprise drop in inflation of the sort we've seen can temporarily raise real wages.

However, in the longer-run, real wages aren't affected by inflation. Instead, they depend upon real things like productivity growth and workers' bargaining power, and these aren't much affected by inflation: at moderate levels of inflation, there's no link between inflation and GDP growth, for example.

This poses a danger - that, in the absence of real supports for real wages, the temporary benefit of lower inflation will be clawed back, in the form of lower future nominal pay rises.

My chart shows the point. It plots the level of real wages (the average wage index divided by the CPI) against the change in real wages in the next four quarters. The relationship is clearly negative. This implies that a surprise rise in real wages because of, say, a surprise drop in inflation leads to lower future pay growth.

This could happen again. It's not just me that thinks so. Here are the minutes of the last MPC meeting:

It was possible that the fall in near-term inflation might become more persistent if it lowered inflation expectations, pay and other cost growth in a way that became self-perpetuating. Inflation had fallen globally and was expected to reach its trough in the United Kingdom in the early part of the year when a large proportion of pay claims were settled. It was therefore possible that the pace of nominal wage growth would be weaker than otherwise.

This is an especial danger because two big determinants of pay growth aren't terribly helpful.

For one thing, productivity is still awful. This week's numbers show that hourse worked rose by 0.5% in September-November. With the NIESR estimating that GDP grew 0.7% then, this gives us productivity growth of just 0.2%.

Secondly, it's not obvious how much bargaining power workers have. Yes, unemployment has fallen which strengthens their hand. But on top of the 1.9m formally out of work there are also 2.3m people outside the labour force who want a job and over 1.3m part-timers who want full-time work. This represents an excess supply of labour equivalent to 13.7% of the working-age population. This might continue to hold down pay growth.

To check this, I ran a quick and crude regression of reage wage growth since 2000 upon lagged unemployment (the official rate), the lagged level of real wages, and productivity growth. Such as equation has an r-squared of 69% since 2000. It tells us that, if productivity rises 1% in the next 12 months then real wages will rise 1.3% (standard error = 1.3pp). This is better than we've seen recently. But it's well below the 2.3% annual growth we saw in the seven years before the recession.

I don't offer this as a definitive forecast. Instead, I offer it as evidence that what matters for real wage growth is not inflation but more fundamental forces. And these aren't yet obviously very favourable.