In a recent column, I argued that the manufacturing sector of the U.S. economy is in relatively good shape, despite the sharp decline in manufacturing employment. I clearly touched a nerve with this column. Not only did I receive a great many emails, but my fellow columnist and mentor Paul Craig Roberts took me to task, as well. I can't respond to everything I heard, but following is a response to the most frequent criticisms.

One common complaint is that U.S. companies are simply reselling goods actually manufactured in China. This is just a misunderstanding of how the gross domestic product is constructed. All imports are subtracted from final sales to calculate GDP. Therefore, imports from China or anywhere else can never raise GPD; they always cause it to be lower than if they were produced domestically. GDP measures only actual production on U.S. soil.

The equation goes like this. In 2002, final sales to domestic purchasers equaled $10,866 billion. You add $3.9 billion for the change in inventories nationwide, add $1,014.9 billion for exports, and then subtract $1,438.5 billion for imports. This leaves a net figure of $10,466.2 billion for GDP. In short, imports reduce GDP and exports increase it.

It is always tempting to think that we can ban imports or tax them in some way and thereby raise domestic output, by forcing consumers and producers to "buy American." The problem is that we import a lot of things we can't produce at all or not enough of domestically, like oil. A lot of imports are industrial supplies and capital goods that are critical inputs into the manufacturing process. Banning them or raising their cost would raise costs for producers, reducing their international competitiveness. It would also invite retaliation by foreign countries. The trade deficit might even rise because exports would fall more than imports fell.

In the end, trade protection has never worked in any country at any time. The long-term effect has always been to impoverish nations that engage in it.

Another criticism I heard is that I used incorrect data to support my point. I looked at total goods production in the U.S., which includes things like mining and agriculture in addition to manufacturing. I did this for two reasons. First, the concern I most often hear from people is that Americans no longer make "things." Therefore, I thought that a broader view of goods output was justified.

Second, data just for manufacturing are harder to come by. Goods data are compiled every quarter, while manufacturing data are available only annually and with a lag. The latest data for manufacturing is for 2001, while we have goods data through the 2nd quarter of this year. Furthermore, manufacturing data after 1987 are incompatible with those before because of certain definitional changes.

Nevertheless, looking at manufacturing alone still makes my point. Since 2001 was a recession year, it is reasonable to compare it to the last recession year in 1991. In nominal (money) terms, manufacturing has fallen from 17.4 percent of GDP to 14.1 percent. But in real (inflation adjusted) terms, it is actually up a little, rising from 16 percent to 16.2 percent. Continued...