My regular column is available to subscribers on www.thetimes.co.uk This is an excerpt. Not to be reproduced without permission.

Those of you who follow these things will know that events have taken a bit of a turn in Newport. Not since Llanelli beat the All Blacks at rugby more than 50 years ago, a few dozen miles away, has there been so much excitement in South Wales.

The kerfuffle in Newport, home of the Office for National Statistics (ONS), has been about the closely watched labour force survey (LFS), and the fact that low response rates have called into question the accuracy of this key set of indicators for policymakers, including those at the Bank of England. They have also created a dilemma for anybody trying to analyse this essential part of the economy, including me.

As things stand, the last full set of LFS data came out more than six weeks ago and covered the period to July, That, in terms of economic data, is if not ancient history, at least very backward looking.

In place of the usual LFS figures, the ONS put together an alternative set, some of it based on experimental data, which is why some economists have been inclined to treat the latest statistics with a large pinch of salt.

Peering through the fog, the new figures appear to confirm a picture of gently rising unemployment and gradually weaker employment, which has yet to get back to pre-pandemic levels. Economic inactivity – the proportion of working-age people who are not part of the labour force, is running at just under 21 per cent, or 8.7 million people, higher than before Covid but lower than its post-pandemic peaks.

The first question to be asked is whether there is anything in these figures to prompt the Bank’s monetary policy committee (MPC) to raise interest rates from the current 5.25 per cent level this week.

The answer to that, taken in the round, is no. Wage growth is still too strong for the Bank’s comfort, but the labour market is loosening, and other activity indicators are weakening. The latest “flash” purchasing managers; index, for this month, dropped to a nine-month low and suggested that the economy is at best limping along. The CBI’s industrial trends survey recorded another fall in manufacturing volumes, and a drop in employment in the sector.

The picture of a softer labour market is confirmed by surveys from the Recruitment and Employment Confederation (REC) and others. It is noticeable that a few months ago many headlines were about worker shortages, which of course persist in some sectors. Now they are outnumbered by stories of job cuts and closures.

You can never say never with the MPC. Last month it surprised the markets with a narrow vote, 5-4, to keep rates on hold. This week it would be a big surprise if the committee were to vote to raise rates. Two surprises in a row would not be unprecedented but would be strange. The Bank should keep official interest rates on hold.

The second question is how much unemployment needs to rise to deliver the lower pay increases consistent with the 2 per cent inflation target.

There is, as you may know, quite a long history to this question. The Phillips’ curve, developed by the New Zealand economist Bill Phillips at the London School of Economics shortly after the Second World War, explained the inverse relationship between unemployment and pay growth. The higher the unemployment rate the lower wage growth, other things being equal.

The Phillips curve has undergone various modifications since, but remains a widely used concept, as does the “natural” rate of unemployment or, to give it its more commonly used but horribly clumsy name, the non-accelerating inflation rate of unemployment (Nairu). All that word salad means is the rate of unemployment consistent with stable inflation.

There is also a bit of history on this topic at the Bank. When, 10 years ago – it seems like longer – Mark Carney took over as Bank governor, he was determined to do things differently. He launched “forward guidance” on interest rates, in contrast to the previous approach under which the MPC looked afresh at rates at each meeting. His version of forward guidance was that the Bank would not consider raising rates until the unemployment rate dropped to 7 per cent, which was not predicted to happen until 2016.

As it turned out, unemployment dropped quite quickly to 7 per cent but the Bank decided it was too soon to raise rates and Carney was on the way to his “unreliable boyfriend” reputation.

The point here is not to revive all that but to remind you that a 7 per cent unemployment rate, equivalent at the time to roughly 2.2 million people out of work, was thought at the time to be close to the Nairu. Drop below it and higher inflation would be the result,

It was too gloomy a view. Though many say that the Bank should indeed have begun raising rates at about that time, unemployment subsequently headed down to less than 4 per cent ahead of the pandemic, without triggering a rise inflation. Inflation averaged less than 2 per cent in 2019.

What about now? Unlike in 2013, we start from a position in which pay is rising by 7 or 8 per cent, double or more the rate consistent with meeting and sticking to the 2 per cent inflation target.

The Bank, in its August monetary policy report, predicted that by the third quarter of 2025, inflation would be back at 2 per cent, alongside earnings growth of 2.5 per cent and an unemployment rate of 4.8 per cent, up from the latest estimated rate of 4.2 per cent but not by much.

The Office for Budget Responsibility (OBR), currently working on its autumn statement forecast, to be published on November 22, was even more sanguine, predicting an unemployment peak of 4.4 per cent, roughly 1.5 million, and for that to be consistent with average earnings growth of less than 2 per cent a year from next year onwards.

I would not expect that forecast to survive the revisions the OBR will undertake during the next three weeks. Both it and the Bank appear to be too optimistic. They suggest the economy will undergo a very soft landing and a near painless adjustment to low inflation. Things that look too good to be true usually are,

A feature of the pandemic was that, while economic inactivity rose, the increase in unemployment was modest, perhaps because of the effects of the furlough scheme.

The previous two recessions, in 1990-92 and 2008-9, resulted in increases in unemployment of between 55 and 65 per cent. That would imply an increase in the unemployment rate from last year’s low point of 3.5 per cent to between 5.5 and 6 per cent. Whether even that will be enough to keep a lid on inflation is the big question.