On October 7, 2016, the Bureau of Labor Statistics estimated that September nonfarm payrolls grew by 156,000. Its household survey showed 354,000 more employed people, seasonally adjusted. Those are not competing totals for the same population. Payroll records count each job separately; the household survey counts a person once and includes groups, such as unincorporated self-employed workers, excluded from the payroll survey. [jobs2016]

The unemployment rate uses another denominator: unemployed people divided by the civilian labor force, not the entire population. September was 5.0 percent, described by BLS as little changed. The payroll baseline was not fixed either. July growth was revised to 252,000 and August to 167,000, leaving their combined gain 7,000 below the earlier estimate. Additional responses and recalculated seasonal factors inform these revisions. [jobs2016]

The Bureau of Economic Analysis supplied a separate measurement lesson in its September 30 report on August. Disposable personal income means income after personal current taxes. It rose 0.2 percent in current dollars but 0.1 percent after inflation. The personal saving rate was 5.7 percent of disposable income, not the percentage of people who saved. These national-account measures cannot describe a particular household budget. [income2016]

For this October 8 retrospective, comparison starts with the measure, its population and its reporting vintage. BLS and BEA provide separate evidence chains about September employment and August income, not a test showing that one caused the other. The BEA release is now marked superseded. Later estimates are not substituted here for what these releases reported at the time. [jobs2016][income2016]

Source documents

  1. Personal Income and Outlays, August 2016https://www.bea.gov/news/2016/personal-income-and-outlays-august-2016
  2. Employment Situation, September 2016https://www.bls.gov/news.release/archives/empsit_10072016.htm