Landfills are the byproduct of a society’s great industrial achievements. However, the placement of landfills has come under scrutiny due to the assertion that they are disproportionally placed in minority communities. Texas is a heavily industrialized state with a wide geographic range and a very diverse population making it an exceptional location to conduct a geographic analysis to answer this question. Using economic, racial, and landfill data from the Census Bureau, I ran an Ordinary Least Squares Analysis to determine whether my explanatory variables did influence the placement of landfills, a Spatial Autocorrelation (Moran’s I) to ensure they are spatially random, and a Geographically Weighted Regression (GWS). The GWS determined that rural minorities did indeed share the brunt of that burden, but rural whites also shared a similar fate. Urban minorities and whites did have a slight overperformance of landfills, but relative to rural Texas, the problem was less severe. These results indicate that there may be a urban-rural divide and the funding that is derived from this divide may be responsible for why landfills are overrepresented in rural Texas as opposed to urban Texas.
Here is a PDF of the project if you would like to download it.
I’m curious about the effects of landfills on the economics of the surrounding region. Of course they are ecologically harmful, which likely is long-term damaging economically, but if rural regions are taking trash from urban regions, aren’t they also earning money by taking people’s trash? I’m not really sure the process in designating and funding landfills, so I wonder how landfills correspond to the economy of their localities.
Hi Henry,
Thanks for your comment. I think this is actually a really interesting perspective to take as it’s not something I considered. It may actually be true that rural places willingly take in waste from urban centers since they have more open space and it would boost their economy in the short-term. I went back and checked my GWS for suburban Texas and noted that they actually have fewer landfills than one would expect. It could totally be true that rural areas willingly take up landfill waste, but that only happens since suburban areas are not open to having landfills near them and urban/suburban centers have to send their waste somewhere that is willing to take it. This may also be a commentary on the general state of waste in the US that we have so much and are not efficient with recycling it that we have this problem in the first place. I could have added a 3rd dimension to this project by also adding in maps from suburban Texas. Thanks for the perspective!
Javier, this is a fascinating topic. When we throw things away, we don’t often think about where “away” is, but it is immensely important to the communities you have identified. Your findings about demographic associations with landfill placement is enlightening, especially for a racially heterogeneous state like Texas. Did you also look at overall population densities in areas with landfills? Even among rural areas or majority-minority communities, I wonder if landfills end up in areas with fewer people or if planners prioritize convenience over safety and place them nearer to people.
Hi Colton,
So, I actually did consider inputting population densities into my analysis, but when trying to run my GWS, that input(explanatory factor) ran into some problems as it made my results biased according to the output report GWS gives me which could have potentially lead me to making bad inferences about the output. I tried to balance out this bias by inputting other explanatory factors, but the model never returned a report without telling me that the population density was biasing the result too much. I ended up utilizing total population of whites and balanced it out with families below poverty level by 100%(minorities are disproportionately poorer than whites) and renters(2/3rds of renters in Texas are minorities) which helped me fix my biasing issue. If I had more time, I would have loved to kept population density since I do agree that that is a very important indicator where landfills do end up being created, but I do not think I had the necessary datasets to look through to find other explanatory factors that could have went in conjunction with population density to not lead to a biased result according to GWS output report. Thanks for your comment!
Excellent initial sounding into this important issue, Javier, good job. I too wonder about how much of the distribution of landfills correlates with population densities–so, when you ran population density on a per-tract basis (people/area of tract), it biased the model? There must be some endogeneity at work there–any ideas about what those endogenous factors might be? Thinking through that might be a path towards a more subtle model. I wonder what kind of R2 you got from this run and what the distribution of your residuals looks like (clustered or random)? These are also clues for next steps. Good going.