Michel Baudin, 12/15/96
The US car industry's public relations message is that it has essentially caught up in productivity with Japanese rivals, and the press is largely chiming in, to the point that it is now assumed true not only for the car industry but for manufacturing in general. Michael Porter, for example, opens an article on strategy in the Nov.-Dec. 96 issue of the Harvard Business Review with a discussion of how, over the past two decades, American manufacturing managers have improved operational effectiveness and now must shift back to strategy. According to all these sources, JIT/lean production is done and it is time to move on. Is that the consensus? Well, not quite. A lone dissenting voice came from outgoing labor secretary Robert Reich, to the effect that the numbers do not bear out the car industry's claims.
What do the numbers show? First, we have to be careful as to which numbers we should be looking at, because the way economists measure productivity is not necessarily relevant on the shop floor. They relate sales dollars to the labor time involved in producing the goods sold. On those terms, the bureau of labor statistics reports that, for all durable goods manufacturing, the output per hour of all persons employed is up 20% since 1992. While such numbers are interesting from a global perspective, they include the influence of such factors as inflation and a weak dollar that are irrelevant to how effectively shop floor resources are deployed.
Specifically in the car industry, we can look at the number of labor hours required per unit built. It is a more relevant measure, but not without drawbacks. It can't be used, for example, to compare two different assembly plants, because their depths of manufacturing may vary. One that outsources engines will use fewer labor hours per car than one that builds them in house, but it doesn't mean it is more productive. With precautions, it can be used for comparison over time of the entire US car industry with itself. We know that, during the past decade, the trend has been towards more outsourcing, and therefore we know that the reductions in labor hours per unit may overestimate but not underestimate shop floor improvements.
We looked up the sales and production numbers for the years 1989 through 1994 in Ward's Automotive Reports and in data from the American Automobile Manufacturers' Association, and matched them against numbers of production workers and of hours worked from the US Bureau of Labor Statistics. The results are shown in the following table:
|
|
|
|
|
|
|
|
|---|---|---|---|---|---|---|
|
New
motor vehicle sales (X1000)
|
15,394
|
14,169
|
12,756
|
12,868
|
13,913
|
15,179
|
|
Sales
of domestic cars and trucks (X1000)
|
11,592
|
10,786
|
9,858
|
10,221
|
11,384
|
12,672
|
|
%
share of transplants
|
10%
|
15%
|
18%
|
18%
|
16%
|
19%
|
|
Average
number of production workers
|
264,325
|
244,750
|
226,042
|
230,600
|
229,933
|
237,717
|
|
Average
number of hours worked
|
43.42
|
42.81
|
42.28
|
41.79
|
42.92
|
46.51
|
|
Labor
hours per car sold
|
44.55
|
43.71
|
43.62
|
42.43
|
39.01
|
39.26
|
|
Productivity
increase over prior year
|
1.88%
|
0.21%
|
2.73%
|
8.07%
|
-0.65%
|
|
|
Cars
sold per production worker
|
44
|
44
|
44
|
44
|
50
|
53
|
|
Number
of hours worked per year
|
1954
|
1926
|
1902
|
1881
|
1931
|
2093
|
Over these six years, the number of labor hours per car sold declined by 12%, or 2% per year, which is hardly evidence of a revolution in production methods, especially when we consider that some of this reduction is due to outsourcing and that the market share of primarily Japanese transplants has doubled in the same period. The Japanese car assembly plants in the US practice JIT/lean production from the beginning, which we estimate makes them about 30% more productive than their mass production competitors. If nothing else had happened, their rise in market share from 10% to 19% by itself would have yielded a 2.7% productivity improvement for the industry as a whole, which leaves barely 9.3% due improvements in other plants. This is the kind of performance that is achieved by routine capital investment, not by conversion to JIT/lean production.
Robert Reich was right. The numbers do not support the claim that the US car industry has massively increased productivity, and its comeback in the past few years is due to other causes, such as the strength of the yen. JIT/lean production is still the exception, not the norm, and most of the implementation work still lies ahead.