Problems with data

Browsing through the links in Economist's View, I find these two articles.
From Entrepreneurship and the Economy:
"So what does entrepreneurship have to do with the recession? If we take what we know today, entrepreneurs and innovation play a vital role in the economy. But can they help us in the great recession? In other words, what policy should we be pursuing to move the unemployment rate below 10 percent and back into the neighborhood of 5 percent? We know that new firms are important. They create most of the net jobs. However, only a small percent, perhaps 4 percent, create almost all of the jobs in any given four-year period. And this seems to hold up in different times, different countries, and different industries.

So how do we forge a policy? Two stories are told out there. First we know that age and size are important variables. And we know that age appears to be more important than size. In other words, we should target firms based on age not size. The two stories out there are one by Zoltan Acs and the other by Carl Shramm. In a highly influential study, Acs found that the average high impact firm was about 20 years old and came in all sizes, small, medium, and large. Schramm, on the other hand, using a Census Bureau study, found that firms less than five years old created almost all of the jobs independent of size. They both cannot be right."

In a different context, from Can You Trust Census Data?By JUSTIN WOLFERS:
"The whole research community is waiting for the Census Bureau to do something about these problems."
Once, when I pointed out some data Kuffirsaid somethinglike "But the ground realities are different". Wise man!