The data? She's hardly econometric...
So there’s this band called Oasis that's on their comeback tour in the UK right now. You might have heard of them. Mainly because UK economists have spent all summer riffing off their lyrics…
But despite all the puns – and promise of a big Oasis-effect on UK inflation – July’s tour dates saw hotel prices surge by… 0.3%. Not exactly a champagne supernova in the CPI... (see I told you it was good!)
So we economists have a new thing to blame… I mean... cite, in our rigorous analysis of forthcoming data points. And fortunately, we don’t have to look very far: data quality issues.
Take the UK, where this week’s retail sales release was pulled because of dodgy numbers. It joins the Labour Force Survey and producer prices in a not-very-exclusive club of data plagued by quality issues.
Time will tell what the latest issue was. But life has got harder for stats agencies around the world – and not just because of funding issues or political pressure. Survey response rates are falling, making key numbers like US non-farm payrolls sketchier. And meet seasonal adjustment – she’s econometric – and an attempt to make data comparable across different times of year, a task made harder by the immense volatility of the pandemic.
There's also the tricky job of filling gaps in the data. Remember that a lot – maybe all – of the US employment gains right now can be traced back to the so-called “birth-death model”, an entirely made-up guesstimate of small business hiring used until more accurate census data comes along. And when it does, that likely means big downward revisions.
Then, more recently, high inflation – and the growing tendency of firms to make big price increases in the first few months of the year – have made it harder to deflate the data. I’ve written before that I think this has distorted UK GDP, where the first half of the year is artificially stronger than the second. That’s surely the next dataset to go under the microscope…
Whatever the reasons, the problem is getting worse, not better. So central banks have just got to (pay)roll with it – and here are five ways they’re likely to react:
First, they're more likely to disagree with one another. And we’re already seeing it. The Bank of England’s August decision failed to reach an initial verdict for the first time ever. The Fed’s July meeting had two governors dissent against the majority verdict, which hasn’t happened for decades. Put simply, data quality – and the volatility it fosters – means more than ever, it’s possible to reach a range of different interpretations from the same set of numbers.
Second, central banks are going to be more reluctant to pre-commit to anything in advance. Admittedly, that’s been true for a while, after officials at both the Fed and, more notably, the ECB, were often forced to go back on their word when inflation took off in 2022.
Third, officials are more likely to focus on surveys. Some might sway, like this week’s Purchasing Managers' Indices in Europe, which hint that the continent’s manufacturing sector is looking brighter. Half the world away, polls show the US jobs market is souring. The closely watched ISM services index shows weaker hiring, while surveys tell us consumers are noticing the cooling jobs backdrop.
By contrast, the hiring surveys here in Britain – having soured ahead of April’s big tax hikes – have started to turn up, hinting that the consistent fall in payroll numbers we’ve seen this year shouldn’t get more dramatic.
Fourth – and this is perhaps more debatable – is whether central bankers are more likely to discount what the jobs data is telling them, which could amplify the asymmetric focus they’re placing on inflation right now. These numbers are generally immune to the quality issues I’ve discussed here.
Next week’s core PCE deflator inflation figures are going to be key, particularly given this week’s Fed minutes revealed that officials think we’re still to see the full effects of tariffs.
Finally, and most importantly, dodgy data means central banks are more likely to be wrong.
If the data officials are looking at is more frequently wrong – or prone to heavy future revisions – the risk of a policy mistake is undoubtedly higher. UK GDP figures released at the time of the financial crisis, for instance, bear little resemblance to the revised numbers we have for the period now.
It’s not the only reason central banks risk ending up behind the curve. I’m a broken record on how fixed-rate lending is slowing down the transmission of rate cuts to the wider economy. And it’s partly why we’re expecting more aggressive easing from the Fed this year. My colleague James Knightley is pencilling three cuts for the remainder of 2025.
Anyway, I’m outta time for this week. Let’s just hope the central banks aren’t too…
THINK Ahead in developed markets
United States (James Knightley):
- Core PCE deflator (Fri): The Fed's preferred measure of inflation is set to match the 0.3% figure we saw from core CPI, in month-on-month terms. It's a close call, though, and the risk is that we get a slightly lower figure. Despite very hot producer price figures, the components that feed directly from that release into core PCE were more mixed.
Eurozone (Bert Colijn)
- Economic sentiment (Thu): Eurozone sentiment will be closely watched next week to see whether it gives August just as shiny a report card as the PMI did last Thursday. The strong manufacturing picture it painted provided hope that the trade war impact has so far been smaller than expected, and confirmation from the European Commission data would cement the view of a better-performing eurozone economy.
THINK Ahead in Central and Eastern Europe
Poland (Adam Antoniak)
- Jul retail sales (Mon): Following a softer June reading, likely driven by a shift toward services spending during the long Corpus Christi weekend, we expect July to mark a return to the retail recovery trend, with renewed demand for durable goods. In 3Q25 private consumption should still be a key force behind GDP growth.
- Aug flash CPI (Fri): We will still work on a more precise nowcast of August CPI inflation as high-frequency data becomes available, but our initial estimate suggests that headline inflation did not change markedly compared to the July reading and was still close to 3%YoY amid broadly stable annual changes in food, fuels and energy prices. In such a scenario, the National Bank of Poland should not hesitate to deliver another 25bp cut next week.
Hungary (Peter Virovacz)
- Rate decision (Tue): Taking everything into account, we still see no scope for the National Bank of Hungary to ease monetary policy in the short term. In line with this view, we expect the central bank to leave the interest rate unchanged at its next rate-setting meeting on 26 August. The base rate will remain at 6.50%, with a +/- 100bp interest rate corridor – a high-conviction call. The main obstacles are the elevated inflation environment (despite government measures) and high household inflation expectations, which are inconsistent with long-term price stability. Moreover, global uncertainty remains, with the forint having the highest beta in the region. Hence, the high-risk premium will continue to be a key factor in preserving market stability and helping to achieve price stability.
- Labour market (Wed/Thu): With a gloomier short-term outlook, companies have continued to rationalise the labour force slowly but surely. However, unfavourable demographics are masking this negative trend, with the supply and demand sides of the labour market shrinking in tandem. Consequently, although the unemployment rate is expected to remain relatively stable, the number of people in employment is predicted to decrease further. With a softer labour market, we expect to see further slowing of wage growth and real wage growth shrinking to a level below the historical average, creating a further obstacle to improving consumer confidence.
Czech Republic (David Havrlant)
- Confidence (Tue): August consumer confidence has likely partially corrected the previous substantial gain, as holidays come to an end and expenses related to the onset of the school year come into view. Still, the mood of households is expected to have remained above the long-term average. Meanwhile, August confidence in manufacturing has likely improved, as the trade agreement between the US and the EU has reduced uncertainty, and overall, the Czech industry appears to be gradually stabilising.
- GDP (Fri): The real GDP growth for 2Q is expected to be confirmed, although there is potential for an upward revision due to strong household spending.
Kazakhstan (Dmitry Dolgin)
- Rate decision (Fri): We expect the National Bank of Kazakhstan to raise the base rate from the current 16.50% at the upcoming monetary policy meeting on 29 August. Several factors support a rate hike, including the nearly two-year high CPI growth of 11.8% YoY as of July, a deteriorating inflationary outlook driven by the recent depreciation of the tenge, and the persistent underperformance of banks' funding growth compared to the acceleration of lending growth despite the announced tightening in the macroprudential measures. Additionally, there has been a renewed growth in households' inflationary expectations recently. While a wait-and-see approach is less likely, it remains a possible alternative, especially given the recent stabilisation of the fiscal deficit. However, even in this scenario, it is still unlikely that a tightening of the monetary stance could be avoided closer to year-end.
