UK Monthly GDP stumbles in more ways than one

This morning the UK received some disappointing economic news as the Office for National Statistics released this.

For July 2023, monthly real gross domestic product (GDP) is estimated to have fallen by 0.5%, with falls in all three main sectors, following growth of 0.5% in June 2023.

It is not the decline per se that is the issue as there had been warning signs, but the size of the decline. Also I have long warned that this is an erratic series and a bad idea and I think that swinging from 0.5% up to 0.5% down rather eloquently makes the point. If that does not then since last December we have gone -0.5% then 0.5% then 0.1%, then -0.3% then 0.2% then -0.1% followed by 0.5% up and then down. Make it make sense!

If we now switch to the ersatz quarterly figures we see a much more stable situation.

Looking over a broader picture, GDP showed 0.2% growth in the three months to July 2023 when compared with the three months to April 2023. Production grew by 0.6% and was the main contributing sector to the three-month growth. Services and construction output also increased in the three months to July, both by 0.1%.

There are at least consistent and hopefully avoid some of the wild swings of the monthly series. Part of that is we will have more data on the earlier months in the series as the latest numbers only have around 60% of the final data.

However the annual comparison does get rather blown up.

Monthly GDP showed no growth in July 2023 compared with the same month last year. For comparison, monthly GDP grew by 0.9% between June 2022 and June 2023.

Part of the issue here is what we might call a Royal problem and perhaps our own fault for having a monarchy.

The Platinum Jubilee in 2022, and the move of the May bank holiday, led to an additional working day in May 2022 and two fewer working days in June 2022. It should also be noted that May 2023, saw one fewer working day as there was an additional bank holiday for the coronation of King Charles III.

These have affected the numbers but in truth there are also other issues with the reliability of the monthly series.

What about July?

We see an immediate issue as we look at our largest economic sector.

Services output fell by 0.5% in July 2023, following growth of 0.2% in June 2023

It very quickly becomes a case of “hello darkness my old friend” as we see this.

The main contributor to the fall in monthly services output was the human health and social work activities sub-sector, which fell by 2.1% in July 2023. This was attributed entirely to a 3.4% fall in the human health activities industry.

Although this time around it was not a consequence of the Covid response.

Industrial action was held in July by NHS senior doctors (two days) and radiographers (two days) for the first time while industrial action by junior doctors increased (five days in July, compared with three in June).

Or at least not directly as we do see the new methodology in play.

 NHS England reported that 65,557 appointments and procedures were cancelled because of the senior doctors strike and 101,977 acute inpatient and outpatient appointments were cancelled because of the industrial action by junior doctors.

The counting of appointments is rather messy, although presumably there would have been a fall under the old GDP system as you do not get paid when on strike.

The next largest faller was information and communication and we do observe quite a wild swing in the main mover and shaker here.

Computer programming, consultancy and related activities was the largest contributing industry, falling by 3.4% in July after three consecutive monthly growths in April, May, and June 2023.

That looks odd but the next one seems more consistent with the changes we looked at in the labour market figures yesterday.

Administrative and support service activities also fell in July 2023, by 1.4%. The largest industry within this sub-sector was employment activities, which fell by 2.3% in July after a growth of 2.1% in June. This industry has had a decrease in monthly output in 9 of the last 12 months.

There is another “hello darkness my old friend” moment. But again it is the strikes.

Education also fell in July 2023, by 1.1%, where the sector saw two days of industrial action in England at the start of the month.

Although this is a bit of a gem.

Please note that education attendance is considered to be constant over the school year, so summer holidays and school leavers did not reduce the estimate of education output in July 2023.

What could go wrong?

Finally there were some positives. But we see some especially wild swings here.

The main offsetting positive contribution was from the arts, entertainment and recreation sub-sector, which grew by 6.6% in July 2023; this was its largest growth since May 2021. Sports activities and amusement and recreation activities grew by 12.4% and creative, arts and entertainment grew by 4.9%.

Production

We see a not dissimilar pattern here.

Production output fell by 0.7% in July 2023, following growth of 1.8% in June 2023 (Figure 4). The largest falling sub-sector was manufacturing, which fell by 0.8% in July 2023.

Manufacturing saw a particularly wide gap with June.

This follows growth of 2.4% in June 2023, the strongest monthly growth since November 2020,

Frankly even allowing for the Bank Holiday impact that looks messy. Indeed our regular swing factor which is pharmaceuticals – for newer readers that industry does not operate on a monthly pattern and is erratic – does not merit a mention in spite of it reducing manufacturing output by 0.16% in July.

Construction

By the standards here this is not a large swing.

Monthly construction output is estimated to have decreased 0.5% in volume terms in July 2023. This follows a 1.6% increase in June 2023,

The Weather

It often gets the blame for weak economic data and it was quoted in the construction release. We can take that wider.

The July ( Met Office) report mentioned “the UK overall rainfall total was 170% of average overall, making this provisionally the wettest July since 2009 and sixth wettest July in the series”. The wet weather was cited as a reason for lower output in retail, as described in our Retail sales bulletin, and also in construction and outdoor accommodation venues.

To be fair the June release did mention that it was probably boosted by the better weather we saw then. So this bit is at least balanced and consistent.

Comment

The monthly GDP series is proving to be what one might call challenging. In impolite company one might call it much worse. I thought I would take a look at what I wrote last month.

Caution is required with taking monthly numbers too literally due to the error range but this time around I think that they are less than the bank holiday impact on the quarterly numbers. I have seen estimates of that being between 0.5% and 0.6% in the past. Personally I think it is not that large but if we halve it then it is material.

Apologies for the clunky English, but I think that holds up on that June allowing for Bank Holidays was perhaps 0.2%. If we add the subtracted element to July it becomes -0.2%. All very back of the envelope but it feels more sensible than what the official series has done.

Switching to the quarterly series then we are in the same position of having growth but not much (0.2%). Since then there has been a change following the Blue Book revisions which in a further muddying of the waters the ONS has ignored today. So again using a back of the envelope calculation the UK economy was 0.8% larger than pre pandemic in July.