An article, appearing in today’s Wall Street Journal (“Strong Christmas at U.S.Stores Didn’t Leave Everyone Merry"), explains the virtues of focusing on government statistics instead of trade data, at least when it comes to retail sales.

Overall, though, holiday sales were stronger than almost anyone predicted as recently as three weeks ago. Even pessimistic retailers had to concede victory over their own expectations that 2004 holiday sales would come up short. The National Retail Federation announced a 5.7% rise in holiday sales 2004, beating both its own forecast of 4.5% growth and the 4.1% rise in 2003.

How did government and the retail industry arrive at such different pictures of holiday retail spending? For one thing, the Commerce Department numbers include sales at gas stations, which rose 21.8%, a category the National Retail Federation doesn’t consider. Even so, the three percentage points separating the Commerce Department’s view from that of the retail industry is an economic chasm that speaks volumes…

The NRF, which began as an association of department stores, counts mainly shopping-mall stores such as department stores and specialty chains… It misses some major kinds of stores: supermarkets, home-improvement stores, drugstores… And it doesn’t capture all online sales.

In contrast, the Commerce Department monthly retail report is based on a survey of 12,000 public and private retail businesses selected to represent the entire U.S, retail sector…

Here's the bottom line, from a source I trust.

The result is the trade group’s sales numbers are increasingly out of step with government data and the full picture of consumer spending… “It’s too weak a link to the true retail sales” trends that occur in the overall economy, says Charles Steindel, senior vice president at the federal Reserve Bank of New York.