Fastfood and Fitness

Part A: Background

South LA's Fast Food Ban

Obesity is a national crisis. About 100 million American adults are clinically obese and the percentage of Blacks and Latino(a)s hovers at around 50% (Centers for Disease Control and Prevention 2018). Obesity is costly. Most of the costs are born by those who are overweight via additional health care costs, and loss of income through lowered productivity and missed work. Our obesity epidemic has enormous indirect costs to society as well. Though estimates vary widely, most agree that those costs easily run into the billions of dollars (Centers for Disease Control and Prevention 2018). In communities where obesity is acute, those direct and indirect costs are magnified. For communities already burdened by crime, inadequate educational opportunities, racism, substance abuse, and other social ailments, the totality of challenges can be nearly insurmountable. South Los Angeles is a community struggling with many challenges, but obesity might be the worst of all because in the long term, it’s probably the most deadly, the most widespread and the most persistent of all threats facing this neighborhood.

In 2008, Los Angeles’ City Council passed an ordinance, labeled by the press as a “fast food ban”, which was held up by local politicians as an effort to strike a blow in the fight to overcome the locally acute obesity crisis. After a few years, the ban was widely condemned by the press as failure of paternalistic governmental overreach. An examination of the ordinance confirms the verdict of the press and the law's critics – it has done nothing to stem obesity in South LA. Some critics argued that the government cannot and should not regulate what people eat - which is a political argument with which you can agree or disagree - but you will still pay the social costs of obesity. Others argued that banning fast food was useless because fast food restaurants probably did not - or definitely did not - factor into the rate of obesity by neighborhood. The argument that local food landscapes are without consequence to the dietary practices of neighborhood residents is one that geographers are well trained to evaluate. We only require appropriate data and effective methodological tools.

Question:
Question:
Question

Part B: Data

In order to evaluate the strength of an argument using statistical tools and data we generally need a hypothesis, and that requires a test capable of using in a hypothesis.

Fitness Data

Proxy Variable: Fifth grade physical fitness results - California Department of Education (download)

5th graders are a good proxy for neighborhood fitness levels because elementary school children rarely travel far to attend school. (unlike high school and middle school children), so they are a good proxy of the fitness levels in their neighborhood
Eliminated most special schools that had too few students or were specialized magnet schools likely to host children from distant neighborhoods
Used 10 years of data, to avoid confusing on especially fit, or unfit cohort of 5th grade students.
Eliminated unusual outliers or schools with very large standard deviations - which indicate unreliable data.
N = x - which is a very large sample

Restaurant Data

County Business Patterns / American Factfinder NAICS codes

Location Quotient

Demographic Data

US Census Bureau / American Factfinder and other datasets derived from the US Census Bureau at the ZIP code level

Part B: Hypothesis and Test

The research question is "Does proximity to an abundance of fast food restaurants have an effect on people's fitness levels?"

Generally, research questions can't be answered directly or completely, so researchers develop tests that allow them to work with the data they have at available.

Proxy variables and proxy questions.

Multi variable problem because fitness levels / obesity does not have a single cause - but with enough data, neighborhood-level trends become evident

Ordinary Least Squares Regression allows researchers to consider the relative causal power of multiple possible variables known to affect the outcome (dependent) variable, which in this case is the aggregate fitness levels of 5th graders over a 10 year span.