The Bureau of Labor Statistics (BLS) September payroll survey finds that employers added a net 114,000 new jobs, continuing the trend of slow employment growth during the recovery. The substantially divergent job growth reported in the household survey and the associated drop in the unemployment rate is inconsistent with other economic indicators and may represent statistical sampling error.
Part of this sluggishness is due to the sharp decline in employment at start-up companies, which has hit record lows. Impending tax increases make it more risky to hire, and excessive regulations make it more expensive to start up new businesses. Congress and the Administration should reduce barriers to starting a business instead of increasing them.
September Payroll Survey Sluggish
Employers added a net 114,000 new jobs in September, with 10,000 of those created in government. The strongest employment growth occurred in the health care sector (+44,000), one sector of the economy that has remained robust during the recession. Employment also rose slightly in transportation and warehousing (+17,000) and financial activities (+13,000) while contracting in manufacturing (–16,000) for the second month in a row. Employment changed little in the economy’s other sectors.
Revisions added 86,000 jobs to the July and August reports. However, virtually all of the increase came from the government—not the private sector.
Employers need to add approximately 125,000 jobs a month to match population growth. In 2011, employers added an average of 153,000 jobs a month. Overall, 2012 job creation has slowed slightly to an average of 146,000 jobs a month but has remained around 100,000 jobs for the past six months. This level of job creation will only slowly reduce unemployment.
Contrary Household Survey
The government conducts two main jobs surveys. One queries employers and asks how many workers they have on payroll. The other polls individuals and asks whether they are working or looking for work. In contrast to the slow jobs growth in the employers’ survey, the household survey reported a jump in employment nearly eight times larger. This curious job growth in turn caused the unemployment rate to fall by 0.3 percentage points to 7.8 percent.
Because these surveys are samples, they have margins of error around the reported figures, just like political polls. If a survey reports a margin of error with a 99 percent probability, it means that one time out of hundred the true figure will be much different from the reported figure. One time out of a hundred for a monthly survey means about once every eight years.
Of the two surveys, the payroll survey has a much smaller statistical margin of error. The real value of the household survey is that it provides information on what non-workers are doing, information that the payroll survey cannot capture. Typically, the two surveys disagree somewhat month-to-month but track one another fairly well over longer periods of time.
What seems to have occurred with the September household survey is the one time in a hundred. The household survey reported an astounding 418,000-person jump in the labor force and an 873,000-worker jump in employment at a time when the economy is stumbling. The last time the household survey showed such a huge jump in employment was in 1983 during the Reagan-era economic boom.
If employers had created almost 900,000 jobs, then GDP should also be expanding rapidly. The payroll survey should also show large job growth, and new unemployment insurance (UI) claims should be dropping sharply. None of these are happening. Second-quarter GDP growth was recently revised down, and new UI claims remain around 370,000 a week.
Today’s economy does not look like the Reagan boom. This month’s household survey is one to set aside until another report comes out next month. There is a good chance that with a new sample, the October report will show a countervailing increase in the unemployment rate.
Start-ups Create Jobs
Policymakers should focus on working to encourage job creation and spur real economic growth. Employment at start-up companies has fallen for five years in a row, reaching unprecedented lows in 2010 and 2011. At the same time, profits at incumbent firms have increased comfortably. What explains this disconnect?
The answer is that most net new jobs come from start-up companies. In fact, existing firms rarely create net new jobs on average. The economy expands primarily through the creation of new firms, not the expansion of existing ones.
Historically, start-ups have continuously created new employment in the U.S. economy, hiring about 15 of every thousand working-age American adults every year for decades. Previous recessions featured big job losses from existing firms. Even in recessions, though, start-up job creation has been strong—until now.
Tim Kane, Chief Economist at the Hudson Institute, finds that start-up hiring has slowed to 10 jobs per thousand adults. In 2010 and 2011 alone, “missing” start-up hiring amounted to 2 million net jobs. With normal start-up hiring, growth in the employment rate would have been three times higher.
Why would start-up employment fall while profits rise? A simple explanation is that the economic and policy environment is too uncertain to take the risk. Forward-looking entrepreneurs have good reason to worry. Unless Congress takes action, taxes will rise by $500 billion on January 1. Many of these taxes will make starting up a new company less profitable. Among other tax increases:
- The top federal income and payroll tax rate will rise from 37.8 percent to 43.4 percent, reducing the income of many small businesses;
- The dividend tax will rise from 15 percent to 43.4 percent; and
- The capital gains tax rate will rise from 15 percent to 23.8 percent.
If President Obama and Congress want to encourage entrepreneurs to start new businesses and hire employees, they should have acted to prevent these tax increases from occurring before Congress recessed in September. The number of small-business owners who identify taxes or excessive regulation as their single largest problem is twice as large (44 percent) as the number who identify poor sales as their principal problem (20 percent).
In a weak labor market with high unemployment, the government should not make this problem even worse. President Obama and Congress have another opportunity in November to minimize the damage from their past inactivity.
Regulation Versus Employment
Another important reason for slow start-up hiring is that fixed costs are rising. In the past four years, regulation has expanded rapidly, raising the cost of compliance.
With new regulations and business requirements in health insurance, small-business finance, environment, energy, and tax compliance, not to mention the ever-expanding reach of state licensure boards, it is expensive to open a business.
All these factors make starting new businesses more expensive, which reduces hiring. In fact, employment at start-up firms is very negatively correlated with the number of federal regulations.
Economy Still Slow
The odd household survey result notwithstanding, the economic data continue to paint a grim picture of an economy muddling along and now at risk of recession as the threat of Taxmageddon draws near.
It is not often that policymakers can take one simple, swift action to improve job growth materially, but this is one of those times. The President and Congress should make preventing Taxmageddon their first task when Congress returns to work in November.
is Senior Policy Analyst in Labor Economics and Salim Furth, PhD, is Senior Policy Analyst in Macroeconomics in the Center for Data Analysis at The Heritage Foundation.
Show references in this report
 For example, the first August report estimated that government employment contracted by 7,000 jobs. The revised report now shows government employment expanded by 45,000. Revisions show that government payrolls expanded by 79,000 more in July and August than the preliminary estimates showed. This accounts for the vast majority of the 86,000 jobs added in the revisions.
 The 90 percent confidence interval for the payroll survey is + or – 100,000 jobs, while the confidence interval for the number of unemployed in the household survey is + or – 280,000 people.
 This excludes the January numbers in each year, which are affected by new population controls introduced because of new census estimates. Household survey employment and unemployment levels both jump substantially in January as the BLS adjusts for a larger population size.
 Salim Furth, PhD, “Why the Slow Economic Recovery?” Heritage Foundation Issue Brief No. 3746, October 3, 2012, http://www.heritage.org/research/reports/2012/10/why-the-slow-economic-recovery.
 John Haltiwanger, Ron S. Jarmin, and Javier Miranda, “Who Creates Jobs? Small vs. Large vs. Young,” working paper, August 2011, http://econweb.umd.edu/~haltiwan/size_age_paper_R&R_Aug_16_2011.pdf (accessed October 1, 2012).
 Tim Kane, “The Collapse of Startups in Job Creation,” Hudson Institute, September 16, 2012, http://www.hudson.org/files/publications/Kane--TheCollapseofStartupsinJobCreation0912web.pdf (accessed October 1, 2012).
 Curtis S. Dubay, “Taxmageddon: Massive Tax Increase Coming in 2013,” Heritage Foundation Issue Brief No. 3558, April 4, 2012, http://www.heritage.org/research/reports/2012/04/taxmageddon-massive-tax-increase-coming-in-2013.
 William C. Dunkelberg and Holly Wade, “NFIB Small Business and Economic Trends,” NFIB Research Foundation, September 2012, http://www.nfib.com/Portals/0/PDF/sbet/sbet201209.pdf.
 James L. Gattuso and Diane Katz, “Red Tape Rising: Obama-Era Regulation at the Three-Year Mark,” Heritage Foundation Backgrounder No. 2663, March 13, 2012, http://www.heritage.org/research/reports/2012/03/red-tape-rising-obama-era-regulation-at-the-three-year-mark; Guinevere Nell, “Obama’s Small-Business Tax Could Average $25,000,” The Heritage Foundation, The Foundry, July 13, 2012, http://blog.heritage.org/2012/07/13/obamas-small-business-tax-could-average-25000/; and Morris M. Kleiner and Alan B. Krueger, “Analyzing the Extent and Influence of Occupational Licensing on the Labor Market,” National Bureau of Economic Research Working Paper No. 14979, May 2009.