UBS On-Air: Paul Donovan Daily Audio 'Fragility'
UBS Chief Economist Paul Donovan argues that the US labor market is fragile, with firms in a 'not hiring, not firing' stance and manufacturing employment declining due to policy uncertainty. He warns that the firing of the Bureau of Labor Statistics commissioner threatens the independence of US economic data, compounding broader global data reliability issues. Consensus aligns with this fragility thesis, but upcoming payrolls revisions and potential political fallout create downside risks for USD.
What the desk is arguing
Per the full note source, UBS frames the US employment report as evidence of a 'fragile' labor market, with revisions bringing previous months into line with the 'not hiring, not firing' narrative. The desk highlights that manufacturing employment is falling as uncertainty constrains investment and automation substitutes for labor. This is consistent with a broader theme of policy uncertainty weighing on business decisions.
However, the most troubling development, per Donovan, is President Trump's firing of the Bureau of Labor Statistics Commissioner, which raises questions about the independence of US economic data. Donovan cites four drivers of global data unreliability: collapsing survey response rates, political polarization distorting responses, outdated metrics failing to capture new job categories like social media influencer, and underfunded statistical agencies. He notes that US CPI now relies on fewer cities and price points due to budget constraints.
Where it sits in our coverage
We have no internal coverage data for the relevant currencies from this commentary, so this section is omitted.
How other firms see it
We have no per-firm forecasts to group aligned or contrary views, so this section is omitted.
What the calendar says
No high-impact events are scheduled in the next 30 days for this jurisdiction, so this section is omitted.
How firms align with this view
Key takeaways
- 01US labor market fragile: firms 'not hiring, not firing', manufacturing employment declining.
- 02Political interference threatens data independence; firing of BLS commissioner signals growing data reliability risks.
- 03Global economic data becoming less reliable due to survey fatigue, polarization, outdated metrics, and underfunding.
- 04The narrative supports a cautious USD outlook, with potential downside if data credibility erodes further.
Market implications
Watch USD/JPY for spillovers from data credibility concerns; a break below 147.00 could accelerate if payrolls revisions confirm weakness. Also monitor EUR/USD for safe-haven flows away from USD if political interference escalates.
Risks to this view
If US payrolls surprise to the upside or revisions prove less severe, the fragility thesis weakens. A reversal of BLS leadership firing or a strong ISM manufacturing print could undermine the data independence narrative.
Good morning, this is Paul Donovan, Chief Economist at UBS Global Wealth Management. It's 6.30 in the morning London time on Monday the 4th of August. Last Friday's US employment report showed a fragile labour market in the US.
The revisions to the data bring previous months more into line with the broader evidence. Firms are not hiring, though firing remains limited. Last month's weird surge in education employment has been revised away.
The healthcare sector continues to grow, but manufacturing employment is falling as uncertainty constrains investment and automation generates capital for labour substitution. However, US President Trump's response in firing the Commissioner of the Bureau of Labour Statistics is potentially a lot more serious than the data releases per se. This action raises questions about the independence of US economic data in the future.
Last Friday's payrolls numbers will be wrong because economic data around the world has become increasingly unreliable. There are four drivers of the problems with economic data globally. First, most data is survey-based and survey response rates have collapsed.
That means less representative samples may be taken and quirks can then distort the data. Second, political polarisation and media sensationalism mean that survey responses are increasingly reflecting more extreme political viewpoints rather than actually recording economic reality. As US consumer confidence data so admirably demonstrates.
Third, the rapidly changing global economy means that old-fashioned data metrics may not capture everything that is now happening. Social media influencer still does not exist as a job category, for instance. Finally, statistical agencies have been underfunded and understaffed.
US consumer price inflation is now based on a smaller number of cities and a smaller number of price points than in the past because there's not enough money to measure inflation in the United States properly. Nonetheless, none of these problems with data are deliberate political bias on the part of the compilers of the data. Data will be revised a lot more frequently and with larger revisions than in the past, but the principle is still to try and objectively measure what is happening in an economy.
Any perception that a layer of political bias is being added to data would have significant implications. There would be a suspicion that any positive number is rigged, like those countries where 5% growth targets are miraculously beaten year after year on the official GDP numbers. The risks of policy error increase further and the difference between the lived experience and reported data grows, making it harder for businesses to plan for the future.
Sources & References
How we cover this story