Economic Geography

Primary data gathering is frequently needed, but many
projects can be undertaken entirely with secondary data
While projects of the type discussed in the preceding section involve primary data
gathering in order to accomplish their purposes, not every project needs such
information. My research has often been involved with data that come entirely
from secondary sources. The wealth of statistical information that is at our finger-
tips today on the Internet is a far cry from the statistical environment we were in
some decades ago. One of my first forays into the use of secondary data for
national scale analysis was the result of a request from Brian Berry. When he was
President of the Association of American Geographers (AAG) he held several
sessions focused on trends in the economy, and he asked me to analyse some data
on trends in regional economies in the United States. Berry was involved in the
conceptualization of the Bureau of Economic Analysis (BEA) Economic Areas
concept, and as BEA began to provide regional data back in the 1970s, the
medium was not in data files as we now know them, but rather paper printouts of
special tabulations. The BEA Economic Areas are a regionalization of the United
States economy that aggregates the approximately 3141 counties into about 
175 metropolitan-area focused ‘core’ areas, surrounded by nonmetropolitan
‘peripheral’ areas. I received a set of data for these regions for the 1965–75 time
190
William B. Beyers

---

period, and spent months coding them onto punch-cards so that I could produce
analyses of trends in income among the BEA economic areas (Beyers 1979).
This particular assignment actually turned out to be a pivotal moment for my
research career for two reasons. First, I discovered how vibrant the service econ-
omy was and how strongly it was associated with regional trends. Second, I real-
ized that non-earnings income (transfer payments and dividends, royalties and
rents) were growing rapidly as sources of personal income. Geographers had not
addressed the role of the latter, in part because regional data were only now
becoming available about these components of the personal income stream. In the
years since undertaking this project, I have repeatedly used the BEA economic area
regionalization to track trends in the United States economy, and have recently
used these data in the context of a minimum requirements model to argue that
all regional growth in the United States in recent years can be explained by trade
in services (Beyers 2005). There are fewer analyses of this type than there should
be, in part because of problems with the disclosure laws that pose difficulties
when aggregating data from the county level to the level of the BEA regional-
ization. These difficulties have thwarted some from undertaking national scale
analyses, as have changes in counting methods (e.g. the shift from the Standard
Industrial Classification (SIC) to the NAICS classification systems).
While many projects can successfully be undertaken with secondary data, it is
also common for there to be a mixture of primary and secondary data use to
make arguments. An example of work of this type is my recent focus on cultural
industries (Beyers 2002). This work was a response to a request for a presenta-
tion to RESER, the European Service Industries Research Network in Bergen,
Norway. While I had done work on arts and cultural organizations as described
above, I had not previously focused nationally on the cultural industries scene. 
In this paper I mixed together a variety of types of data, ranging from analyses of
the personal consumption expenditures accounts that showed rising demand for
spending on cultural services (in real $), to data from studies I had undertaken of
recreation, arts, and sports. I tried to use these data to contextualize the relative
importance of these activities in the national economy, and used BEA data to try
to identify something about the geography of consumption of these activities. 
I reported data on the structure of income and expenditures, as well as regarding
the unequal incomes earned by professional sports figures and people working in
the arts. I also brought various results from economic impact studies together to
show the relative contribution of components of these sectors to the regional
economic base. I think that this hybrid approach worked well to touch upon a
number of key attributes of a relatively understudied part of our economy.