The February 2009 percentage reflected the steep decline in the homebuilding industry once the recession (as normal people define it) began in July 2008. February 2010 is high because of a mini-rebound tied to first-time homebuyers’ desires to take a tax credit which expired two months later. February 2011′s low percentage is tied to the industry’s bottom, which occurred at about that time. The high 2012 and 2013 figures are the result of respective slumps which occurred during the previous summer and fall.

In broad historical context, more thanks to luck than anything else, the bureau’s seasonally adjusted annual rate of 917,000 looks reasonable. But does it accurately reflect the economy as we know it today?

If you believe that calendar year 2013 and subsequent years will be repeats of 2011 and 2012, when early-month hopes were dashed in the spring and summer by higher gas prices, ongoing regulatory uncertainty, and heavy-handedness, and gathering domestic and worldwide fiscal storms, your answer would be “no.” In that case, you would only want to consider how February 2012 and February 2013 compared to the rest of their previous 12 months.

If you did that, as seen above, you would calculate a two-year “seasonally adjusted” annual rate of 798,000 units (62,400 actual starts divided by 7.82 percent), a result which is a huge 13 percent below the government’s official number.

Instead of believing that the housing market is on the verge of breaking out, your take would be that the market for new homes is still treading water. Who’s right? No one can possibly know, and that’s precisely the point.

In housing, employment, unemployment claims, retail sales, production, and so many other areas, seasonal patterns which had been mostly intact for decades or had slowly evolved have been broken. Those who blindly rely on seasonally adjusted calculations — yes, I’m talking to you, business reporters in the establishment press — act as if none of this historic breakage has occurred.

In this uncertain economy, conscientious analysts and economists should be digging deep into the raw — not seasonally adjusted — numbers and attempting to interpret them. I see little evidence that this is happening. Their clients, and the public, are not well-served.