Investors assessing the dismal labor market might be worried to learn of the even bleaker picture painted Friday by the Labor Department's household employment survey, which some economists consider a more accurate measure than payrolls. In February, the household survey showed that employment plummeted by 265,000. What's more, the jobless rate, which held steady at 5.6%, would have climbed to 5.8% if 392,000 workers had not simply dropped out of the labor force. "It's not a good report," said Paul Kasriel, chief economist at Northern Trust. "I don't think there's any positive spin that you can put on it." A number of economists have said the household survey is a better gauge of employment than the payroll survey because it takes into account self-employed workers. This survey has shown that 1.3 million jobs have been created over the past year. In comparison, the payroll data show that just 122,000 jobs have been created. Last month, however, both surveys painted a similarly weak picture of the labor market. According to the payroll data, just 21,000 jobs were added in February, well below the consensus estimate of 130,000. Some economists said the weather had an impact, noting that construction jobs fell 24,000. But January's payroll additions were revised down to 97,000 from 112,000, and December's were lowered to 8,000 from 16,000. The Bureau of Labor Statistics calls about 60,000 homes to compile the household survey but contacts 400,000 businesses to establish the level of payrolls. Because the payroll sample is so much larger, covering millions of workers, many economists feel this set of data is more accurate, but the debate about which survey is more reliable has raged on. Brian Wesbury, chief economist at Griffin, Kubik, Stephens & Thompson, believes the payroll data have been understating the level of job growth, saying these statistics aren't keeping pace with the dynamic nature of the economy and all the new business formation.