A recent Washington Post editorial launched another "occasional series about inequality" with a misplaced question mark -- "A Rising Tide?"
 
"In the 25 years from 1980 from 2004," the editorial asserted, "the wages of typical workers fell slightly after accounting for inflation. So, too, did wages for the 50 percent of the workforce that earned less than the typical, or median, employee."
 
As we learned from similar efforts in the past, such statistics tend to be patched together to fit a preconceived theme, with minimal regard for accuracy.

A previous Washington Post "occasional series about the changes roiling the middle of the American workforce" was launched in September 2004 with a three-page feature on "The Vanishing Middle-Class Job." Readers were then warned to be alarmed because: "In 1967 nearly a quarter (22.3 percent) of households made between $35,000 and $49,999 in inflation-adjusted terms. But that share was down to 15 percent by 2003."

However, a closer look at those same figures showed the percentage of households earning more than $50,000 was up to 44.1 percent in 2003 from only 24.9 percent in 1967. Too many people making too much money was a novel way to describe a crisis.

After such an embarrassing debut for their last "occasional series" on muddling middle income, you might think they'd be more careful this time. Apparently not.

This year's inaugural editorial referred to a  median wage -- the point at which half earn more and half earn less. According to the Bureau of Labor Statistics (BLS), the "usual median earnings of full-time wage and salary workers" amounted to $262 a week in 1980 and $638 in 2004.

Translated into 2000 dollars, using the deflator for personal consumption, the median weekly wage rose to $589.40 in 2004 from $503.09 in 1980 -- an increase of 17.2 percent. Real wages have always fallen when oil prices surged (such as 1981 and 1990), yet the real median wage for 2005 nonetheless remained higher than any previous year except 2004.

There were three recessions between 1980 and 2004, so comparing figures for the start and end of that period cannot show what happened when. The real median wage in 1991 was $510.68, for example, so the increase by 2004 was 15.4 percent, or 1.2 percent a year. That sounds OK to me, but nobody knows whether it was slow or fast. We can't compare that 1.2 percent annual gain with the past because this data series began in 1979.