The annual revisions to the Commerce Department’s manufacturing shipments were disappointing for the US commercial printing industry.  Every five years, there is an Economic Census of business establishments, and in intervening years there is a survey of business, the Annual Survey of Manufactures (ASM), which captures shipments and other data. Businesses (and governments) need market data more often. For that the Commerce Department conducts Manufacturers’ Shipments, Inventories, and Orders (M3) Survey orders, inventories, and shipments, which are watched carefully by the financial markets. The best known report from this survey is the durable goods orders report released near the end of each month.

Think of it this way: in a government data Olympics, the Economic Census would be the gold, ASM would be silver, and the M3 Survey would be bronze. The latter two are surveys, and as such they are subject to forms of survey bias. The first is a census; while it does involve surveys, the methodology is more complex and the number of participants is much higher. The M3 Survey is the smaller of the two surveys, so the effect of statistical bias issues are larger. Those problems involve the “representativeness” of the respondents and the number of respondents. The government has a way to deal with that, and that comes from the tax system, especially data reported about payroll employment and the tax system. The official statement with the data revisions was “Correcting monthly data for late receipts, reclassifications of reported data, and revisions to previously reported data.” All of these sources contribute to the revisions of GDP that are published annually at the end of July.

The corrections and refinements of the M3 Survey take time to be done. While the survey data may be monthly, it can take years to to settle what the final numbers are through tax and other data.