Today’s economic data was a bit confusing (unemployment at 5.9%, new jobs not what was expected at 57,000), but the economy added yet another 26,000 new businesses in October, and that’s more than 1,000 new ones every business day. But one thing was quite clear and unequivocal in the manufacturing shipments data: printing is still having problems. Unadjusted shipments for October 2003 were $700 million less than October 2002, when the general economy was in its deep lackluster funk. Though unadjusted shipments were up from September by $71 million, the seasonally adjusted data were down by $150 million.

Shipments for the industry are down, at -2.7% for the year, according to this report. Needless to say, this is extremely disappointing and does not bode well for hopes that a growing economy, even a fast-growing one, can get the aggregate business of print growing again. Of course, well-run individual businesses can do quite well in this environment as weaker competitors struggle. In fact, some surveys of individual print businesses may indicate a growing market while industry-wide data series like these do not. I discuss this in detail below. Below is the graph of my special data series showing the seasonally- and inflation-adjusted 12-month moving total of industry shipments. About half of the decline occurred before 9/11/01.

Official report: http://www.census.gov/indicator/www/m3/prel/pdf/s-i-o.pdf Why is this data series important? This data series is important because it shows the direction of total industry volume. By examining it we can understand how much of the economic recovery is spilling over into our industry. We already know anecdotally that we are not getting much of a rebound, based on the financial reporting of most public print businesses and the continuing downsizing and reorganizing actions taking place within industry suppliers. What is of most concern is that the data below show the disconnect between GDP growth and the growth of our industry.

We long assumed that they were still connected, but as the chart below shows, that connection was lost some time ago. It’s important to keep a few things in mind when looking at these data and comparing them with data about printers’ business situations that are derived from surveys. Some recent surveys imply a more robust print economy than is actually there. It’s not that the surveys are wrong, and it’s not that they are misleading. Here’s what’s happening. The data series presented above is industry shipments, for the total print and print services industry. It is quite different than surveys that ask printers what their sales increase was or what it will be. My cautions about "survivor bias" should always be kept in mind. Basically, most surveys about printers’ business conditions should be interpreted as "the printers who are still in business expect THEIR OWN SALES to go up by x%."

To arrive at an industry forecast, these survey data must be reconciled with the demographics of the industry. That is, you must take another mathematical step. For example, using very round (and meaningless) numbers: --- Assume that there are 1000 printers doing $1000 each in sales. - This would make industry sales volume $1,000,000 ($1,000 x 1000 shops) - Then assume that 100 printers disappear, leaving 900. - A survey says that the remaining 900 printers expect their sales to go up by an average of 10%. - Then industry volume would be 900 x $1100, or $990,000.

Therefore, the survey said that sales would increase by 10%, but industry volume would be $10,000 less, a 1% decrease in industry volume. Where did the 10% increase come from? The volume from the out-of-business printers. The remaining shops did experience an increase in the sales, but not enough to increase overall industry volume. So just keep this in mind when judging and using published industry data derived from surveys. You’ve got to do ALL the math when determining the size, scope, and direction of the business as a whole.