New York: Big business, we keep being told, has been so hampered by regulatory uncertainty over the past few years, it has been reluctant to hire workers.

So it is surprising to read the results of a little-known survey from the Bureau of Labour Statistics: Very large businesses, it turns out, have been expanding their domestic workforces relatively rapidly.

If, since January 2011, businesses of all sizes had hired at the same rate as those with 5,000 or more employees, we would have almost four million more jobs today.

The BLS data show the overall weakness in the US labour market reflects slow employment growth among smaller businesses, following a recession that hit them especially hard.

The data come from the bureau's Job Openings and Labour Turnover Survey (JOLTS), which is commonly used to examine trends in rates of hiring and job loss. While the monthly payroll numbers that receive so much attention show the net change in employment — the difference between the number of workers hired and the number leaving their jobs — the JOLTS data provide crucial information about why employment is rising or falling. The JOLTS data are also useful in that they are broken down by business size. As the economists Alan Krueger, now the chairman of President Barack Obama's Council of Economic Advisers, and Sarah Charnes, at the Treasury Department, wrote last year, the JOLTS data are "the most timely government source of information on employment trends by establishment size."

Business establishments with 5,000 or more employees expanded their workforces by an average of 0.4 percentage point a month from January 2011 to February 2012, the most recent data show. In the same period, in contrast, businesses with 10 to 50 employees expanded by only 0.1 percentage point a month — and those with nine or fewer employees actually reduced their net employment.

Trends

These data are based on establishments rather than entire companies. Since large firms could include many smaller operations, it may also be useful to examine trends for establishments of more than and less than 250 workers to get a better sense of how employment trends are varying by size.

Since January 2011, establishments with more than 250 workers have expanded their employment by an average of 0.25 percentage point per month, compared with an average of 0.15 percentage point for smaller ones.

If all establishments had expanded at the same rate as those of more than 250 workers, we would have almost two million more jobs today. So depending on how you do the calculation, if, since the start of last year, employment growth at smaller businesses had matched that at larger ones, we would now have two million to four million more jobs.

To be sure, the JOLTS data by establishment size are not an official BLS data series — rather, they are "experimental." These types of surveys involve sampling problems along with difficult conceptual issues.

Contrasting pattern

Private-sector indexes show a contrasting pattern. The ADP National Employment Report, for example, which is based on statistics from the payrolls that Automatic Data Processing manages, has shown more rapid employment growth over the past year for businesses with fewer than 500 employees than for those with more than 500. Similarly, Intuit's small-business employment index shows stronger small-business growth than the JOLTS data do. Those private measures, however, don't categorise very large businesses separately — and the JOLTS data suggest the very large businesses are expanding more rapidly than any others.