Putting Quantitative Data in Context

October 2007

By John Rees, Junior Project Manager

AngelouEconomics

At last month's annual International Economic Development Council (IEDC) conference, the increasing influence of technology on economic development was readily apparent. Among the venders present were a host of burgeoning firms offering state-of-the art software applications for data manipulation and display. The products and services on display promised revolutionary increases in data analysis to today's economic development professional.

Advances in data collection and examination cannot be dismissed. Before a community enters the imagination of a site selector, the community's website has been visited and mined for information. Often the initial data presentation determines whether or not a community is considered for further consideration. Aggressive communities find success in presenting a comprehensive quantitative portrait that can be easily digested by outside observers. Transparency is vital.

As data has ubiquitous, though, the risk of relying exclusively on quantitative information becomes increasingly tempting. Too often a comprehensive list of data points substitutes for a true understanding of the business dynamics within a given environment. Numerical information soon becomes gospel, leaving subjective judgment and qualitative factors unexamined.

Despite the increasingly sophisticated nature of tools to gather and analyze quantitative information, the collection of qualitative information remains vital in providing a context with which to examine quantitative information. Although quantitative information can provide an initial understanding of a community, such information has its limitations. Examples of the failure of quantitative to fully capture the essence of a situation are both plentiful and powerful.

When the City of Nicholls Georgia recently undertook a comprehensive planning process, initial figures suggested that the community was undergoing explosive population growth. For decades, the rural community suffered from anemic growth. During the 1990s, for example, Nicholls added just 5 residents. According to the 2000 Census, the city's total population was 1,008. In 2006, however, the American Community Survey placed Nicholls population at 2,706.

The census numbers caught nearly everyone by surprise; within Nicholls there was little evidence of any growth, much less on the scale suggested by the Census Bureau. New housing construction was virtually non-existent. School enrollment was flat. Job growth was modest. How was it possible for a community to nearly triple its population within 6 years without any outward signs of growth?

In 2005, Corrections Corporation of America opened a prison in Nicholls. Equipped with 1,640 beds, the prison reached capacity within months of completion. Importantly, the Census counts prisoners as residents in the communities in which they are incarcerated. While such calculations do little to impact large urban areas, they can greatly distort the statistics of small towns. Today, for example, only a minority of Nicholls ‘residents' are free citizens.

Recognizing the unusual demographic characteristics is key to understanding Nicholls. Georgia requires that the comprehensive plans of all communities include a projection of future population growth. A statistical regression of future growth based on the numbers published by the census produces a population estimate of over 13,000 in Nicholls by 2025. If prisoners are removed from the equation, the 2025 population estimate is 3,000, a significantly more realistic picture of future growth. Such discrepancies have profound implications for Nicholls, especially in the area of service provision.

While the Nicholls population numbers were technically correct, many other communities have directly denounced Census estimates as fundamentally flawed. Challenges to population estimates continue to rise each year. In 2001, only 3 communities challenged population estimates; by 2005 this number had grown to 45. For many communities, the decision to challenge census estimates is based on fiscal considerations, as many states allocate resources based on population counts. Other communities are more concerned with the marketing implications of potential population declines. For older urban areas that have experienced significant declines in past years, projecting an image of current prosperity is extremely difficult when the census suggests that the community is still hemorrhaging residents. Regardless of motivation, most cities undertake census challenges based on the belief that official quantitative numbers do not reflect reality.

Perhaps no city in America has been as aggressive in highlighting the methodological limitations of the census population estimates as has the City of St. Louis . For each of the past five years, St. Louis has demanded changes in its official population estimates. In 2003, the Census estimated that the city's population had dropped by a few thousand residents over the previous year. Civic leaders, however, remained unconvinced by the numbers. Old industrial buildings were increasingly being redeveloped as hip condominiums aimed at a growing young professional class. Chic cafes, bars, and coffeehouses were reinvigorating once empty streets. On the ground, all evidence pointed to a resurgence in city life.

City officials, dubious about the numbers published by the census, decided to collect their own quantitative data. Specifically, St. Louis challenged the methodology behind the numbers. Census population estimates are based in part on new housing construction and the demolition of older residential units. Importantly, the Census Bureau uses an obsolescence schedule to calculate the destruction of aging buildings. For example, every year the Census calculates that approximately .03 percent of housing built prior to 1940 is destroyed. In older cities that are undergoing revitalization, however, such housing is more likely to be renovated. Ultimately, the collection of exhaustive quantitative information helped the city prove triumphant in its many challenges.

In response to the census methodology, St. Louis now keeps comprehensive records of all housing units constructed and destroyed. Such data has been paramount in St. Louis's continued success in challenging census estimates. Instead of a declining population, today's revised estimates show that St. Louis enjoys sustained growth. Although the increases are modest, they do reflect the success of city officials in bringing residents back to the city.

The revised numbers of St. Louis provide evidence not only of the fallibility of Census population estimates at the local level, but at the state level as well. For example, if the census accepts a city's population challenge and increases the local population by 10,000, the bureau does not change the national population count. Instead, the census simply subtracts a total 10,000 residents from states that do not challenge population estimates; since 2000, more than 100,000 residents have been removed from the population of California. While such a number may represent a small percentage of California's overall population, the changes highlight the potentially subjective nature of quantitative information.

During my final day at the IEDC conference, I returned to the trade show. One vendor demonstrated a promising software tool for site selectors. After entering numerous criteria into the program, I was given a list of cities that satisfied all the demands of my fictitious client. The information included sophisticated maps identifying the location of my ideal locations, accompanied by comprehensive data on each community. Although my client was imaginary, I asked which of the community could provide the necessary environmental permits within 90 days. The vendor paused a moment before replying, “You'd have to talk with the relevant city officials to find that out.” Indeed. For all the promising possibilities presenting by today's technology, old-fashioned discussions remain vital to the site selection process.