Statistical Errors and Their Impact on Central Banks
The past month has been challenging for statisticians, with both the US Bureau of Labor Statistics (BLS) and the UK’s Office for National Statistics (ONS) admitting to errors in their data. The BLS acknowledged that it had messed up the weights for its Current Population Survey, affecting employment, unemployment, and labor market participation figures. Similarly, the ONS came clean about overstating April’s inflation statistics due to incorrect weighting of vehicle taxation data.
Bogus Labor Snapshot
The BLS has been facing concerns about the quality of its data, including a weighting error in the employment figures that was deemed minimal but will be corrected in the May data. Additionally, staff shortages have reduced the quality of the official US CPI inflation data, causing a suspension of price collection in some locations and more data than usual being imputed. The BLS has also stopped collecting and publishing many items in its producer price index, which will hurt the "taco trade" due to the suspension of tortilla manufacturing price indices.
The non-farm payroll data, a flagship labor market product, has been found to have poor preliminary estimates. Since 2023, non-farm payroll growth has almost always been revised down, with revisions of almost 50,000 a month between the latest data and the first estimate. This is a significant issue, as the Federal Reserve closely watches these figures and makes decisions based on them.
Obviously Not Sound
The UK’s ONS is facing its own set of problems, including a review into its culture and leadership and the sudden resignation of its chief statistician. The agency has been battling with a broken Labour Force Survey, which prevents the Bank of England from knowing what is happening to participation in the labor market. The ONS has also been boasting about clear improvements in the data and survey response rates, but the numbers tell a different story, with a 35% response rate for the first wave of interviews falling to less than 14% by the fifth.
Dirty Laundry
The UK and US statistical agencies should be praised for publicly airing their mistakes. Economic statistics are getting harder to collect, and it’s essential to acknowledge errors and work towards improving data quality. Other countries, like China, regularly delete uncomfortable data, while EU statistics can give misleading results. For instance, Ireland’s quarterly growth rate was first estimated at 3.2% but then jumped to 9.7% after revision, which is not reflective of genuine economic activity.
What Can We Do?
To navigate the challenging world of data, it’s essential to:
- Not rely solely on a single statistic to make important decisions
- Use multiple sources of data to extract a common signal
- Seek to generate unbiased indicators using modern econometrics
- Not skimp on funding statistical agencies and change laws to make it easier to use administrative data
Governments and central banks must prioritize data quality to make informed decisions. By acknowledging errors and working towards improvement, we can build a more reliable and trustworthy statistical system.
Conclusion
In conclusion, the recent errors in statistical data have highlighted the importance of reliable and accurate information for central banks and policymakers. By understanding the limitations and challenges of data collection, we can work towards creating a more robust and trustworthy statistical system. It’s crucial to prioritize data quality, use multiple sources, and generate unbiased indicators to make informed decisions. Ultimately, this will help us better navigate the complex world of economics and make more accurate predictions about future trends and patterns.
What I’ve Been Reading and Watching
The US CPI inflation figures for May were benign, with little sign of tariffs driving prices higher. However, there were early signs of aggressive pricing behavior in a few areas, such as banana prices, major appliances, and toys. This cautionary signal should encourage the Fed to continue monitoring the situation closely.
A Chart That Matters
A chart showing the monthly annualized change in banana prices, which shot up in May, highlights the need for caution. While this is far from an inflationary surge in overall prices, it’s essential to keep a close eye on pricing behavior in various industries to anticipate potential trends and patterns.
