The annual gap in household income may reach $60 billion. Are signs of a recession in the United States becoming increasingly apparent?
2025-09-12 17:39:58
The US dollar index rebounded 0.18% to around 97.70 during the European session on Friday (September 12). US stock indices hit record highs, driven by the surge in US tech stocks, providing a short-term boost to the dollar. On Thursday evening, US Department of Labor data showed that the August Consumer Price Index (CPI) rose roughly in line with market expectations, paving the way for the Federal Reserve to cut interest rates by 25 basis points next week and at least one more rate cut before the end of the year. While inflation concerns have been temporarily alleviated, initial jobless claims last week increased at the fastest pace in four years, refocusing market attention on the US employment crisis.
We might as well rethink the revised non-farm employment data released by the U.S. Department of Labor on Tuesday. The total non-farm employment in the first 12 months of March 2025 was revised down by 911,000 from the previous estimate.
A baseline revision of this magnitude reinforces the view that the labor market is cooling faster. This revision shows that non-farm payrolls were overestimated by nearly 1 million jobs (the largest revision since the global financial crisis) and essentially wipes out the significant job growth in official statistics for 2024, which also means that household income is weaker than previously reported.
Assuming the average worker earns $36 an hour and works 35 hours a week (excluding non-salary benefits such as healthcare and retirement contributions), the overestimation of employment data will result in an estimated $60 billion annual shortfall in household income, representing approximately 0.5% of total household income. This shortfall is further exacerbated when social welfare benefits and employer contributions to retirement expenses are taken into account.
The U.S. Bureau of Economic Analysis (BEA) revise its gross domestic product (GDP) estimates based on updated data from the Bureau of Labor Statistics (BLS). Because BLS employment data feeds directly into personal income calculations, which in turn influence national income statistics, any downward revision in employment data will inevitably impact GDP.
In theory, national income should be aligned with economic demand, as represented by GDP. Therefore, if income data is revised downward, GDP will also face a downward revision of at least 0.5%, unless household savings data is also revised downward.
It is important to note that the non-farm payroll data from the U.S. Bureau of Labor Statistics (BLS) comes from a monthly survey that tracks monthly changes in employment, hours, and income across all non-farm sectors, including private enterprises and government departments.
The survey sample covers approximately 122,000 companies and more than 666,000 independent workplaces, covering nearly every core sector of the U.S. economy. Employers are required to report the number of employees in the payroll period that includes the 12th of each month, as well as submit salary and work-hour data.
The U.S. Bureau of Labor Statistics (BLS) then releases preliminary estimates, revises them within the next two months, and conducts a more comprehensive benchmark revision in January of each year based on administrative records of unemployment insurance claims.
The September revised data released this time is a preliminary revised version, and the final revision results will be announced in January next year.
Since the outbreak of the COVID-19 pandemic, the integrity of the Current Employment Statistics survey (CES, the official name of the non-farm payroll survey) has continued to come under pressure.
The survey's response rate—historically averaging around 60%—dropped sharply in the early months of the pandemic. Data collection was hampered by the closure of numerous businesses, particularly those in the service sector, or the shift to remote work or other inability to contact them.
This weakened the representativeness of the survey sample, which has never returned to pre-pandemic response rates—data for hard-hit industries is underreported, and the probability of small businesses dropping out of the survey sample has increased significantly. This decline in sample coverage has injected more disturbances into the monthly estimates, leading to unusually large revisions, such as the 2024 employment data.
In addition, in several cases, methodological difficulties and classification misjudgments (such as counting employees on leave as active employees) further distorted the core data.
Although the U.S. Bureau of Labor Statistics (BLS) has adjusted its statistical processes to account for such anomalies, the reliability of initial non-farm payroll data has declined significantly in the post-epidemic era.
Meanwhile, on Wednesday, the U.S. Department of Labor’s internal watchdog said it had launched an investigation into how the Bureau of Labor Statistics (BLS) collects employment and inflation data.
This means that the market needs to interpret such data with caution and combine it with other economic indicators (such as the Institute for Supply Management (ISM) and the Conference Board's employment survey) - these surveys actually show that the employment situation has deteriorated significantly.
The long-term weakening of the US dollar index also suggests that the fundamentals of the United States are changing.
Currently, the US dollar index is barely holding above the 5-day moving average and the lower edge of the box, but is under pressure from the 10-, 20-, and 30-day moving averages, all of which are pointing downward and tend to continue to diverge, suppressing the rebound of the US dollar index. 97.40 is a key support level, the point where the US dollar index quickly rebounded after breaking the bottom on September 9th.
KDJ remains below 50 and MACD is below the zero axis, suggesting that the rebound of the US dollar index may end at any time and needs to retest the support below.
Thursday's decline in the US dollar index also failed to reach this level. The nearest resistance level is the September 10 high, also around 98.00. However, constrained by the weakening US labor market and the Federal Reserve's potential imminent easing cycle, the US dollar index may continue its downward trend.
(Daily chart of the US dollar index, source: Yihuitong)
At 17:15 Beijing time, the US dollar index was at 97.70.
We might as well rethink the revised non-farm employment data released by the U.S. Department of Labor on Tuesday. The total non-farm employment in the first 12 months of March 2025 was revised down by 911,000 from the previous estimate.
A baseline revision of this magnitude reinforces the view that the labor market is cooling faster. This revision shows that non-farm payrolls were overestimated by nearly 1 million jobs (the largest revision since the global financial crisis) and essentially wipes out the significant job growth in official statistics for 2024, which also means that household income is weaker than previously reported.
The U.S. job market may be more serious than expected
Assuming the average worker earns $36 an hour and works 35 hours a week (excluding non-salary benefits such as healthcare and retirement contributions), the overestimation of employment data will result in an estimated $60 billion annual shortfall in household income, representing approximately 0.5% of total household income. This shortfall is further exacerbated when social welfare benefits and employer contributions to retirement expenses are taken into account.
The U.S. Bureau of Economic Analysis (BEA) revise its gross domestic product (GDP) estimates based on updated data from the Bureau of Labor Statistics (BLS). Because BLS employment data feeds directly into personal income calculations, which in turn influence national income statistics, any downward revision in employment data will inevitably impact GDP.
In theory, national income should be aligned with economic demand, as represented by GDP. Therefore, if income data is revised downward, GDP will also face a downward revision of at least 0.5%, unless household savings data is also revised downward.
Overview of the Bureau of Labor Statistics' statistical methods
It is important to note that the non-farm payroll data from the U.S. Bureau of Labor Statistics (BLS) comes from a monthly survey that tracks monthly changes in employment, hours, and income across all non-farm sectors, including private enterprises and government departments.
The survey sample covers approximately 122,000 companies and more than 666,000 independent workplaces, covering nearly every core sector of the U.S. economy. Employers are required to report the number of employees in the payroll period that includes the 12th of each month, as well as submit salary and work-hour data.
The U.S. Bureau of Labor Statistics (BLS) then releases preliminary estimates, revises them within the next two months, and conducts a more comprehensive benchmark revision in January of each year based on administrative records of unemployment insurance claims.
New circumstances make it more difficult for the Labor Bureau to compile data
The September revised data released this time is a preliminary revised version, and the final revision results will be announced in January next year.
Since the outbreak of the COVID-19 pandemic, the integrity of the Current Employment Statistics survey (CES, the official name of the non-farm payroll survey) has continued to come under pressure.
The survey's response rate—historically averaging around 60%—dropped sharply in the early months of the pandemic. Data collection was hampered by the closure of numerous businesses, particularly those in the service sector, or the shift to remote work or other inability to contact them.
This weakened the representativeness of the survey sample, which has never returned to pre-pandemic response rates—data for hard-hit industries is underreported, and the probability of small businesses dropping out of the survey sample has increased significantly. This decline in sample coverage has injected more disturbances into the monthly estimates, leading to unusually large revisions, such as the 2024 employment data.
Core data, the market needs to interpret such data with caution
In addition, in several cases, methodological difficulties and classification misjudgments (such as counting employees on leave as active employees) further distorted the core data.
Although the U.S. Bureau of Labor Statistics (BLS) has adjusted its statistical processes to account for such anomalies, the reliability of initial non-farm payroll data has declined significantly in the post-epidemic era.
Meanwhile, on Wednesday, the U.S. Department of Labor’s internal watchdog said it had launched an investigation into how the Bureau of Labor Statistics (BLS) collects employment and inflation data.
This means that the market needs to interpret such data with caution and combine it with other economic indicators (such as the Institute for Supply Management (ISM) and the Conference Board's employment survey) - these surveys actually show that the employment situation has deteriorated significantly.
Technical Analysis
The long-term weakening of the US dollar index also suggests that the fundamentals of the United States are changing.
Currently, the US dollar index is barely holding above the 5-day moving average and the lower edge of the box, but is under pressure from the 10-, 20-, and 30-day moving averages, all of which are pointing downward and tend to continue to diverge, suppressing the rebound of the US dollar index. 97.40 is a key support level, the point where the US dollar index quickly rebounded after breaking the bottom on September 9th.
KDJ remains below 50 and MACD is below the zero axis, suggesting that the rebound of the US dollar index may end at any time and needs to retest the support below.
Thursday's decline in the US dollar index also failed to reach this level. The nearest resistance level is the September 10 high, also around 98.00. However, constrained by the weakening US labor market and the Federal Reserve's potential imminent easing cycle, the US dollar index may continue its downward trend.
(Daily chart of the US dollar index, source: Yihuitong)
At 17:15 Beijing time, the US dollar index was at 97.70.
- Risk Warning and Disclaimer
- The market involves risk, and trading may not be suitable for all investors. This article is for reference only and does not constitute personal investment advice, nor does it take into account certain users’ specific investment objectives, financial situation, or other needs. Any investment decisions made based on this information are at your own risk.
