Goldman Sachs (GS) and a few other Wall Street firms forecast real GDP growth of less than 2% this quarter. And yet, the unemployment rate in January dropped to 8.3% – the lowest level in three years. The decline goes against Okun’s Law, which economists have historically relied on to forecast what the job market might look like given how quickly (or slowly) the economy is growing. As a rule of thumb, Okun holds that year-on-year economic growth of 2 percentage points above the trend — widely considered 2.5% — is needed to lower unemployment by one point. And vice versa.

Since the Great Recession, the unemployment rate has defied the law.

James Pethokoukis of the American Enterprise Institute, a Washington DC-based think tank, has laid out three instances: In 2009, the unemployment rate edged to 10% following a 3.5% drop in GDP. But under Okun, unemployment should have risen higher to 10.4%.

MORE: The 7 elections that will reshape the global economy

At the end of 2010, the unemployment rate fell to 9.4% from 9.9% the previous year. But given that the 3% rise in GDP was barely above trend, the jobless rate should have stayed flat. And in 2011, when GDP rose a point below trend to 1.7%, Okun would have predicted that unemployment would rise to 9.9%. However, it actually fell to 8.5% from 9.4%.

All this has made many wonder if the economy is doing better than what the data currently shows or if unemployment seems artificially low.

It could be that today’s GDP statistics are wrong. The economy might actually be growing much faster than we think, which wouldn’t be too surprising since it’s not unusual for growth statistics to get revised years later as economic data comes in. In a research note to clients on Monday, JP Morgan (JPM) economist James Glassman pointed to the 2008-2009 recession in which GDP was significantly revised downward last year.

“At the time employment trends were much weaker than the impression left by the real GDP trends,” he noted. “That was three years after the fact. With the economy now recovering, there is a high probability that preliminary estimates of national output eventually will be revised up.”

Certainly that could happen, but that still doesn’t capture the whole jobless picture.

The unemployment rate is also influenced by the labor participation rate – that is, the percentage of working-age persons who are employed as well as unemployed and searching for work. While labor participation has been stabilizing recently, it has declined considerably over the years. And at 64%, the rate is two perecentage points lower than its pre-recession level. As Fortune pointed out last week, the drop might have less to do with discouraged workers giving up their job hunt (as economists widely believe), but also the flux of aging baby boomers retiring and leaving the labor pool altogether.

MORE: Unemployment: Leveling off or just a retirement influx?

Admittedly, it’s one of the most uncertain times to retire. And there’s evidence that older workers are working longer as retirement plans and pension benefits become harder to rely on. But as Barclays Capital (BCS) economist Dean Maki has pointed out, lower and middle-income retirees rely more on Social Security benefits as their sole source of income.

What’s more, while some older workers might hold off on retirement, others might forgo a paycheck earlier for disability benefits. In a research note to clients in February, JP Morgan economist Micheal Feroli noted that recipients of disability benefits have accelerated since the recession. The trend was highlighted in Fortune in 2010, but Feroli offers some updated figures that reveal how the rise in disability insurance has driven down the unemployment rate. In a research note to clients in February 2012, he noted that since the onset of the downturn, the growth of disability benefits may account for up to one-fourth of the two-percentage point decline in the labor force participation rate.

“Once individuals go on the disability rolls, they rarely go off,” Feroli writes. “All this suggests that – even after holding demographics constant – the labor force participation rate may only partly recover as the labor market picks up.”

We probably will have to wait a few more years to know what’s really going on. For now, we know that the decline in unemployment doesn’t just mean more people are working today.