Published: December 23, 2010
In 2010, journalist Dante Chinni and statistician James Gimpel published Our Patchwork Nation. Seeking to move beyond the simplistic blue and red states paradigm, they examined the nation’s 3,141 counties in terms of race, education, income, religion, and politics to get a flavor of local perspectives. They identified 12 different community types based on “common experiences and shared realities” which they breezily named boomtowns, campus and careers, emptying nests, Evangelical epicenters, immigration nation, industrial metropolis, military bastions, minority central, monied burbs, Mormon outposts, service-worker centers, and tractor country. See the Patchwork Nation website for a detailed explanation of each community type.
Selecting twelve community types was not an arbitrary decision. The authors tried to represent each county as a particular mixture of fundamental community types. Although an extremely accurate decomposition might seem like it would require a large number of community types, sophisticated mathematics saves the day. A technique called principle component analysis (PCA), allows the community types to be ordered in terms of how much the addition of each extra community type helps to explain the county to county variance seen in the collective racial, education, income, political, and religious data. As noted in the appendix of Patchwork Nation, their analysis showed that ten to fourteen community types, but most commonly twelve, explained 70-80% of the variance and thus represent a good tradeoff between utility and accuracy. Each county is then scored on how well it matches each community type and for simplicity assigned the community type it scores highest on.
Community names are derived by examining the trends driving each principle component. Categories like Monied Burbs, Boom Towns, and Service Worker Centers communities are primarily drawn along economic lines. The Immigration Nation and Minority Central communities are distinguished primarily by racial composition. Not all religious difference are significant on a large scale; no community types explicitly call out Protestant denominations, Catholics, or Jews. However Evangelicals and rural Mormons do stand out in the big picture.
After reading the book, I was curious if the number of bridge players was significantly different among the community types. After a bit of programming I boiled the results down to the table below.
|Community Type||Active Players
in full population
in full population
in 45+ y.o. pop
in $60,000+ pop
|Campus and Careers||417||114||109||107||1377||113||3413||108|
|Service Worker Centers||251||69||63||61||736||61||2788||88|
|United States (all)||365||100||102||100||1215||100||3170||100|
On average, there are 365 bridge players per million people in the United States and considerable variation among the community types. The Monied Burbs, which includes San Diego county, have 45% more bridge players per capita. San Diego county has 505 players per million, slightly below average for its community type. Campus and Careers communities have the second highest proportion of players. I didn’t expect this result given how few college students are playing bridge now; however this concern appears to be offset by the overall elevated education levels and probably the tendency of intellectuals to retire in university town. Conversely, I expected the Empty Nests community to rank slightly higher. Rural Mormons just don’t play bridge — if they’re going to send missionaries our way perhaps we should send some of our own their way.
I limited the analysis to active players, defined as ACBL members who have played in at least one sanctioned game during the twelve month period examined. Restricting further, the second pair of blue and yellow shaded columns shows the number of regular players, defined as those ACBL members who have played at least one sanctioned game during each of the 12 months in the period examined. The pattern for regular players follows that for active players, more easily seen by comparing the first two yellow columns where the numbers in the corresponding blue column have been scaled so that the United States average is always 100. But Tractor Country has relatively fewer regular players. This may reflect the low population density of Tractor Country; maintaining a regular game may be a challenge.
The average age of ACBL members is nearly 70. Since the age distributions vary in each community type, it is interesting to consider the number of bridge players per capita, restricted to the population 45 years old or older. Above this age, the average number of bridge players is 1215 per million. But the yellow column shows that this age correction does not have a major impact on the ranking of the community types.
Bridge players are also fairly well off. So instead of normalizing by age, I tried normalizing by household income, specifically to that part of the population with an annual household income of $60,000 or more. Now the communities start to look alike – except for the rural Mormons who still don’t play bridge even if they are well off. Interestingly the number of players shoots up significantly in the Evangelical Epicenters.
Income, or perhaps wealth, is important in keeping the game alive. Bridge isn’t that expensive at the club level but one must have some leisure time and some disposable income.
Is disposable income and leisure time really the determining factor? Bridge tends to attract smart people and smart people tend to earn more money. Perhaps the income correlation is just a byproduct of the community’s intelligence. This is probably true at some level though a quick check of literature for studies correlating intelligence with income or wealth, indicate only a modest correlation.
Analysis details: Zip codes for ACBL members in the United States in 2009 were mapped to county names and the county names were mapped to FIPS codes, a unique numeric identifier assigned by the U.S. Census. The mapping failed in ~0.2% of the cases, usually due to stale zip codes. The analysis included Washington D.C. as a pseudo county but excluded the U.S. territories.
Active and regular players were determined by examining data from October 2009 to October 2010. Because I was missing one of the monthly masterpoint reporting files, a regular player is not actually someone who has played in 12 of the last 12 months but rather someone who has played in all 12 of the last 13 months for which I have activity data.
After filtering for active and regular players, the FIPS codes were used to join the community type designation and U.S. census data kindly provided by the Dr. Gimpel. The population data is from 2006. For simplicity, I treated each county as belonging to its primary community type. Since the Patchwork Nation data scores each country against each community type, it is possible to do a little better. Some day … when I have a bit more time. If you must know the answer, download my count of bridge players per county and take it from there. In considering the number of bridge players per capita older than 45 or with a household income $60,000 or higher, it would be ideal to remove bridge players not meeting those criterion. Unfortunately, I do not have the data to do this. For the age cut, very few players would be removed; the income cut is a bit more problematic but the present analysis is probably not a horrible distortion.
Download data. Zip archive includes data for the table above and the count of “active” and “regular” players in each county in both tab delimited text format and Excel format. Tabular data is to two decimal places and includes additional data for regular players not shown above. The Patchwork Nation data can be requested from Dr. Gimpel, jgimpelgvpt.umd.edu.
More on Mormons and CardsFrom the 1st edition of Mormon Doctrine:
CARD PLAYINGHmm… I’m not sure about the “lying, cheating spirit” but bridge tables have indeed been “the scene of many quarrels” and even a murder or two. And yes, bridge does become a “ruling passion” for some. Does a double squeeze constitute a “devious and dark way” to win?
See Apostasy, Gambling, Recreation. President Joseph F. Smith has stated the position of the Church with reference to card playing in these words: "Card playing is an excessive pleasure; it is intoxicating and, therefore, in the nature of a vice. It is generally the companion of the cigarette and the wine glass, and the latter lead to the poolroom and the gambling hall... Few indulge frequently in card playing in whose lives it does not become a ruling passion... A deck of cards in the hands of a faithful servant of God is a satire upon religion... Those who thus indulge are not fit to administer in sacred ordinances... The bishops are charged with the responsibility for the evil, and it is their duty to see that it is abolished... No man who is addicted to card playing shall be called to act as a ward teacher; such men cannot be consistent advocates of that which they do not themselves practice.
“The card table has been the scene of too many quarrels, the birthplace of too many hatreds, the occasion of too many murders to admit one word of justification for the lying, cheating spirit which it too often engenders in the hearts of its devotees...
“Card playing is a game of chance, and because it is a game of chance it has its tricks. It encourages tricks; its devotees measure their success at the table by their ability through devious and dark ways to win. It creates a spirit of cunning and devises hidden and secret means, and cheating at cards is almost synonymous with playing at cards.” (Gospel Doctrine, 5th ed., pp. 328-332.)
Members of the Church should not belong to bridge or other type of card clubs, and they should neither play cards nor have them in their homes. By cards is meant, of course, the spotted face cards used by gamblers. To the extent that church members play cards they are out of harmony with their inspired leaders. Innocent non-gambling games played with other types of cards, except for the waste of time in many instances, are not objectionable.