4Alpha Research Analyst: Kamiu
In today's global economic field, the importance of employment data to global macroeconomic policy makers and trading markets is self-evident. As an important indicator of economic development, the non-farm payroll data in the United States has always been highly regarded. However, there has long been a voice of doubt in the market: why do the US employment data diverge from the trend of CPI, and why is there a significant difference between household survey and establishment survey data? This divergence has led some people to doubt the accuracy of the non-farm payroll data released by the US Department of Labor, believing that there may be errors, or even systematic overestimation, especially as non-farm data has been abnormally high since 2024, and the non-farm data for July 2024 has dropped significantly beyond expectations, further raising systematic doubts about the non-farm data.
Next, we will explore the reasons behind this phenomenon and the potential impact it may have on market analysis and policy making.
I. Why has US employment data long been suspected of being inaccurate, or even systematically overestimated?
The non-farm payroll data released by the US Department of Labor (BLS) every month, including employment numbers, unemployment rates, etc., has always been regarded as one of the most important macroeconomic indicators. The addition of non-farm employment reflects the number of new jobs in the non-agricultural sector in the United States, including all industries outside of the government sector, such as manufacturing, services, construction, etc. This data helps to understand the speed of expansion in the US job market and the degree of tightness in the labor market. The unemployment rate refers to the proportion of the labor force in an unemployed state to the total labor force within a certain period of time. It is another important indicator for measuring the health of the economy, reflecting the degree of idle labor in the labor market. Average hourly earnings reflect the income level of American workers and are an important indicator for measuring consumer purchasing power and potential inflationary pressures.
Non-farm data has significant impact on financial markets, government policy making, and economic forecasts. Investors, economists, and policy makers closely monitor this report to assess the trend of the US economy and make corresponding investments and decisions. The performance of non-farm data often affects the monetary policy of the Federal Reserve, thereby affecting global financial markets. However, in recent years, an increasing number of views believe that US employment data is inaccurate and may be systematically overestimated, mainly due to the following reasons:
The differences between non-farm data from different sources are becoming increasingly significant (details will be discussed below), and the lack of robustness in the data is increasingly prominent, leading to questioning the credibility of non-farm employment data.
There are certain potential contradictions between different macroeconomic data. Under the recent significant decline in CPI data, the employment market still shows a continuous moderate growth trend, as specifically compared below:
January 2024:
CPI: According to data from the US Bureau of Labor Statistics, the CPI in January decreased by 0.1% month-on-month and increased by 6.4% year-on-year.
Non-farm payroll data: The number of new non-farm jobs in January was 517,000, and the unemployment rate remained at 3.4%.
February 2024:
CPI: The CPI in February remained unchanged month-on-month and increased by 6.0% year-on-year.
Non-farm payroll data: The number of new non-farm jobs in February was 311,000, and the unemployment rate slightly decreased to 3.3%.
March 2024:
CPI: The CPI in March decreased by 0.2% month-on-month and increased by 5.2% year-on-year.
Non-farm payroll data: The number of new non-farm jobs in March was 235,000, and the unemployment rate remained unchanged.
April 2024:
CPI: The CPI in April decreased by 0.4% month-on-month and increased by 4.9% year-on-year.
Non-farm payroll data: The number of new non-farm jobs in April was 213,000, and the unemployment rate slightly increased to 3.4%.
May 2024:
CPI: The CPI in May decreased by 0.3% month-on-month and increased by 4.0% year-on-year.
Non-farm payroll data: The number of new non-farm jobs in May was 184,000, and the unemployment rate remained at 3.4%.
June 2024:
CPI: The CPI in June decreased by 0.2% month-on-month and increased by 3.2% year-on-year.
Non-farm payroll data: The number of new non-farm jobs in June was 176,000, and the unemployment rate slightly decreased to 3.3%.
The above data depicts a somewhat strange scenario, that is, in the first half of 2024, the CPI in the United States showed a monthly decline, but the non-farm employment numbers continued to rise moderately, showing strong resilience, which does not conform to the naive predictions made by observers based on the Phillips curve. Although the Phillips curve has been repeatedly proven to have limited fitting and predictive ability for actual situations in history, and its specific elasticity is also a long-standing debate in the macroeconomic community, the continuous deviation of the data from the Phillips curve over a long time scale since 2023 still raises doubts about the data itself (this article temporarily sets aside the discussion of the statistical caliber of CPI).
The sub-data contained in the non-farm data are contradictory to each other. For example, the non-farm employment data for May 2024, which is widely regarded as the most bizarre in the past decade, recorded a significant increase in employment, but the unemployment rate increased significantly without a significant increase in the labor force, forming an inexplicable self-contradiction (of course, the May non-farm payroll data has been significantly revised downward in June, but this has further intensified the market and commentary's doubts about the reliability of the initial data).
Since 2024, non-farm employment data has been revised downward multiple times. Since 2023, the non-farm employment data published by the US Bureau of Labor Statistics has been revised downward multiple times. For example, the non-farm data for May 2024 showed an addition of 272,000 jobs, far exceeding the market's expected 185,000 jobs, but the multiple downward revisions of the non-farm data before this have led to doubts about the accuracy of this data. The Philadelphia Fed even suggested that the additional job data in 2023 may have been overestimated by as much as 800,000.
Non-farm employment data contradicts other employment survey data and has consistently exceeded collective economist forecasts. Quarterly Employment and Wages Survey (QCEW) and the ADP National Employment Report for private businesses in the United States have long shown signs of a cooling labor market in recent months, but the non-farm data has consistently shown unexpected resilience in the US employment situation. It is generally believed that non-farm employment data relatively does not distinguish between formal and informal employment, while QCEW and others are more biased towards formal employment statistics, with limited statistics on informal and part-time employment.
II. Brief introduction to how non-farm employment data is specifically calculated
BLS compiles non-farm data based on a series of detailed surveys and statistical methods. The following are the key steps and methods for calculating non-farm employment data:
Sample survey: BLS collects data through the Current Population Survey (CPS) for household surveys and the Current Employment Statistics (CES) for establishment surveys. The household survey is mainly used to calculate the unemployment rate and labor force participation rate, while the establishment survey is used to calculate the number of new jobs and average hourly earnings.
Industry classification: Non-farm employment data categorizes employment into different industry categories, such as manufacturing, construction, services, etc., in order to analyze the employment situation in each industry in more detail.
Data adjustments: This mainly includes seasonal adjustments and Birth/Death (B/D) adjustments:
To ensure the accuracy of the data, BLS conducts seasonal adjustments to eliminate the impact of seasonal factors on employment data. Specifically, BLS first analyzes historical data to identify and quantify seasonal patterns. Seasonal patterns refer to the fluctuations in employment data caused by regular or predictable factors (such as holidays, weather changes, school holidays, etc.) during specific time periods. Secondly, BLS uses the S-ARIMA time series analysis method to fit model parameters that make the residuals white noise using historical data, and performs seasonal differencing on the original data to eliminate seasonal fluctuations.
As CES surveys cannot capture the employment changes in newly established and closed businesses in real time, BLS uses the Birth/Death Adjustment model to estimate these changes in order to more accurately reflect the actual employment market. This includes the Birth Model, which estimates the jobs created by newly established businesses. This model is based on historical data and considers the growth trends and macroeconomic conditions in different industries to predict the contribution of new businesses to the job market; and the Death Model, which estimates the jobs lost due to closed businesses. This model also relies on historical data to analyze the frequency and patterns of business closures, as well as the impact of macroeconomic conditions on business survival.
III. Conclusion: Is US employment data intentionally overestimated?
The author believes that at the level of being questioned, there is a striking similarity between CPI and non-farm data. These two monthly data with significant macroeconomic significance have always been repeatedly questioned by the market as to whether they are manipulated to meet the needs of the incumbent political figures in the United States for support and votes, thereby questioning the independence of the Federal Reserve. Of course, the author cannot completely rule out the possibility of this conspiracy theory being true, but still believes that the various anomalies and inconsistencies in non-farm data in recent years are more likely due to the interrelated reasons of outdated statistical methods, structural changes in the US economy after the pandemic, and the increasing rate of illegal immigration.
- Outdated Statistical Methods
As described below, the operating mode of the US economy may have undergone structural changes, but the seasonal adjustments and B/D adjustments in CES data highly rely on historical data patterns, which may lead to significant deviations, with the B/D adjustment being the most criticized.
According to the data, in May, 231,000 of the new non-farm jobs came from the B/D model, which is an estimate of the jobs created by new businesses. These job positions were not actually counted as being created, but were assumed to exist and directly included in the data. Since April 2023, the B/D model has added 1.9 million jobs, accounting for 56% of all new jobs during the same period. This means that over the past year, more than half of the "employment growth" has come from adjustments, leading to the majority of market views pointing to the B/D model as the main culprit for the "unreasonable" May 2024 non-farm data, as shown in the figure below. In recent years, the percentage difference between CES and CPS results has been increasing, and it is also considered as strong evidence that the CES sampling and statistical adjustment methods have severely failed.
- Structural Changes in the US Economy after the Pandemic
Before and after the COVID-19 public health event, there has been a significant increase in the proportion of informal work and a rapid decline in the employment willingness of young people, which has continued to this day. There is currently no particularly strong explanation for this phenomenon, and some believe that the increase in the proportion of informal work and the decrease in employment willingness may be due to the long-term sequelae of COVID-19 reducing the overall labor capacity at the population level, but there is no consensus. In any case, it can be confirmed that the increase in part-time work will greatly increase the difficulty of non-farm employment statistics, as the non-farm data is obtained through sampling surveys, and the same person engaging in multiple part-time jobs will inevitably lead to an overestimation of job positions compared to the actual situation, and eliminating this noise will lead to disproportionate survey costs. At the same time, a large number of eligible people exiting the labor force (the denominator of the unemployment rate) will also lead to distorted statistics of the unemployment rate and the number of job additions.
- Ineffective Border Control, Increasing Rate of Illegal Immigration
This is closely related to the aforementioned structural changes in the economy, as the probability of illegal immigrants engaging in informal work is significantly higher. Additionally, the employment of illegal immigrants can also lead to potential sampling bias.
BLS's non-farm employment data is based on CES sampling surveys. If the sample does not adequately represent the employment situation of illegal immigrants, the survey results may deviate from the actual situation. For example, if the CES survey's sample (employers as sampling units) covers more large enterprises that are inclined to hire legal workers and overlooks small or underground enterprises where illegal immigrants are more likely to work, then the employment data is highly likely to be overestimated.
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
