There's been a bit of banter on the back of the GDP data talking down manufacturing, and suggestions that the trend is towards renewed recession.
Indeed as Adam Lent points out, the broader industrial sector (including utilities and extraction) has declined two quarters in a row - meeting economists' rule-of-thumb definition of recession.
Yet, we urge caution on talking down manufactuing specifically, in part because a weak April and a quirk of the statistics masks a rather strong May and June for manufacturing.
Here's what today's quarterly data suggests about the monthly production figures for manufaturing:
Monthly manufacturing output, 2006=100
So the rather sharp dip of 1.6% in April was more than made up by 2.1% growth in May and June. So, absolutely, we see that April was a tough month, not just because of the royal wedding and the bank holidays, but because that's when the Japanese tsunami hit UK production the hardest.
But there's also a statistical quirk at play.
Even though manufacturing output ended the quarter higher than where it started, output officially fell by 0.3% across the quarter. That's because the quarterly data are produced through averages. A weak April dragged down the q2 average, even though output ended the quarter higher than any point this year. (The June index number is implied by today's quareterly data). So compared to the quarterly average in q1, q2 is weak, even though output was higher in 2 of 3 months in q2 than in any monh in q1.
And if you go back through the data, you'll see that manufacturing output is at its highest since activity fell off a cliff in November 2008.
Monthly manufacturing data for 2011, 2006=100
But don't take this as economists having it both ways or spin from vested interests. Manufacturing outpu remains 9% below its pre-recession peak and was fairly flat in the first half of 2011.
That's because manufacturers in the UK have been grappling with some fairly fierce headwinds: rising commodity prices, skills shortages, a lack of finance for SMEs and weakness and uncertainty in key export markets have all weighed on the sector this year. Nor do they look like abating in the second half of 2011.
So while we're not complacent that a strong manufacturing rebound in 2010 has given way to a more modest 2011, we do think May and June offer encouraging signs that shouldn't be played down.
NB: If you clicked through Duncan Weldon's blog on the latest IoP data, you'll see he's talking about 3-month on 3-month changes, while this post flags up the monthly figures. Typically, it's better to look at the 3-month data as Duncan does because it gets rid of monthly volatility to provide a truer picture of trends. I've focused on the monthly numbers here, not because they help prove my point, but because we need to understand why April is bad, and how manufacturing responded afterwards. In this case, the quarterly averages simply wipe away any of the useful information coming from the data.