Rhode Island’s state government employs 23% fewer workers now than it did two decades ago, while local government payrolls have grown 11% over the same period, according to data prepared for WPRI.com by the Rhode Island Department of Labor and Training.

In May 1990, state government employment in Rhode Island totaled 21,200; as of last month, that number was down to 16,400, a decrease of 4,800 jobs over 21 years. (DLT’s records begin in 1990.)

The decline at the state level was partly offset by the growing payrolls of cities and towns. Municipal governments added 3,400 jobs from May 1990 to May 2011, raising their total employment from 31,700 to 35,100. The number of local government jobs peaked at 39,500 back in 2003.

Another way to look at the figures is to compare them with Rhode Island’s total population. The number of state employees was equal to 1.5% of the population in 2010, down from 2.1% in 1990, while the number of municipal employees ticked up from 3.2% to 3.4% over the same period.

Rhode Island’s situation apparently isn’t unique. State employment levels are down across the country, according to Chris Mauro, a municipal-bond strategist at RBC Capital Markets.

“These headcount reductions are the best indication, in real time no less, that the states are making the tough, but painful decisions required to bring their budgets into balance and, as a result, should be well positioned when the economy ultimately improves,” he wrote in a recent report quoted by the WSJ’s MarketBeat blog.

Total private- and public-sector employment in Rhode Island is up 1.8% compared with 21 years ago, from 459,000 jobs in May 1990 to 467,100 last month. State government’s share of all jobs fell from 4.6% to 3.5%, while local government’s share rose from 6.9% to 7.5%.

Excluding Census years, federal employment in Rhode Island has been relatively stable over the past two decades, totaling 10,200 last month. That’s down only 800 from the peak of 11,000 jobs in 1997.

One caveat: May 1990 was two months before the start of the 1990-91 recession, so the data likely captures a high point of state employment (though we don’t know that for sure). That means we’re comparing the peak of a business cycle to a less robust point at present, as my e-mailers would rightly point out.

Related: Honey, I shrunk the state work force – considerably (March 10)