Given the heated debates over the local benefits, or the direct opposite, of having a Walmart come to town, plus the sheer centrality of this once humble little Southern enterprise in America's retailing, it was inevitable that economic science would take on The Walmart Question. The results to date shed more light and less heat than is usually the case when discussing the national impacts of Walmart's outsourcing in Asia or the impacts on communities when Walmart comes to town.

Our understanding on this last point, the very way we ask questions about local impacts, has been deepened by new research at the Federal government's Northeast Regional Center for Rural Development, which focuses on New York, New Jersey, Pennsylvania, and the New England States. The NRCRD's present geographical mandate is wider than its name implies: it includes the small towns and small cities that are outside the big, sprawling urban-suburban regions — places like the towns along both sides of the Shawangunk Ridge and in the Catskills.

"Walmart and Social Capital" is the subject, and the title, of a disquieting article published in the American Journal of Agricultural Economics in 2006. It was written jointly by Penn State's Stephan Goetz, the director of the NRCRD, and Anil Rupasingha, the director of the Small Farms Program at Cornell. The AJAE is a principal international reference point for research on both the food sector and regional development outside big cities. The fact that this journal published disquieting results on a heated controversy is witness to the quality of the research. Sloppy reasoning and weak data are not at home in the AJAE.

Before getting into the nitty-gritty of what makes Goetz and Rupasingha's findings so disquieting, I would like to present a consumers' information sticker. I am not an unalterable foe of Walmart. As I write this article, I'm dressed from head to toe in sturdy cotton clothes purchased at Walmart, with the exception of my shirt, which is from Sam's Club. That may change.

The Wall Street Journal, not exactly a left-wing rag, ran pieces on Walmart's extraction of up to an hour's worth of unpaid work from their employees. My to-do list includes an item to follow up on that story. I'll make my decision whether to continue shopping at Walmart on the basis of whether the company cleans up its labor practices. And also on a basic market criterion: a low-priced item is not a bargain if you also get lower quality.

Back to the chase. Moving a stage beyond the substantial body of research on Walmart's impacts on local small businesses and labor markets, Goetz and Rupasingha targeted the social networks, organizations, and civic capacity to act and solve problems. They were well-positioned to take this on. Along with their colleague David Freshwater at the University of Kentucky, Goetz and Rupasingha went to great lengths to translate the general and competing concepts of social capital into measurements based on concrete, reliable data for every county in the United States.

Why did they go to such great lengths? Because it is increasingly apparent to economists that social capital is a basic factor in whether towns and regions prosper or not — and should have been long ago.

Different aspects of local social capital were measured concretely by things like the number of voluntary associations per 10,000 people in a count; nonprofit organizations per 10,000 people; church adherence and attendance; voter turnout, and filling out census forms. These numbers catch the degree to which a community has, literally, its act together.

Then all the counties were sorted into three Walmart categories: 1) one or more local Walmarts existed in 1990; 2) new Walmarts were established, or additional stores built, in the 1990s; 3) counties without Walmarts. Next they folded in a generous basket of other social and economic characteristics like levels of education or income that are known to have major impacts on social capital. Then the data was sifted with sophisticated math to make sure that any positive or negative changes in social capital were due to the presence or non-presence of Walmart. For example, lower average income typically clobbers social capital. The math compares low income counties with and without a Walmart to see if the big box store has an independent effect on social capital.

Here's what they found:

Walmart generally locates its new stores in counties at the low end of the social capital scale. The level of social capital drops further yet after Walmart arrives.

When a new Walmart came to town during the 1990s, the number of voluntary associations and charitable organizations declined.

Voter participation in the 2000 presidential election declined.

Church adherence declined.

There were also some paradoxes: in counties where Walmart was already present in 1990, church adherence went up. And, in presumably a show of some community spirit, the percentage of people filling out the census form increased. Otherwise social capital stayed low. You can fill in the blanks as to why.

Thank God that human beings are a curious species. If that wasn't the case, people like me would be out of business.