Monday, October 1, 2012

GIS and Local Politics: How GIS is Helpful in Analyzing Voting Patterns


In its early Years GIS software was used purely to create basic maps, but as time has gone on the possibilities for what it can do have grown exponentially. GIS has even found a major place in the world of political science. In their 2002 article A GIS-based spatial analysis on neighborhood effects and voter turn-out: a case study in College Station, Texas Danile Z. Sui and Peter J. Hugill uncover the base of what can be done in the field of Political Science using GIS. They hypothesize that neighborhood effects manifest themselves in different ways according to the spatial distribution of voter turn-out, and they use GIS to prove it.
The study area is Brian/College station a city with a population of 160,000 people in 2006, and the home to Texas A&M a university with 42,000 students. The complex system of city government in place causes many local issues. This in turn makes College station a perfect location for this experiment to occur in. This experiment looked at three major issues:
1)      In 1995: a $10 Million Bond for the acquisition of 3.5 acres of land for municipal service facilities, and to reconstruct, improve, and extend existing streets, to construct and improve sidewalks, traffic signals, and necessary drainage.
2)      In 1997: the construction of a city convention center.
3)      In 1999: an issue over an overused residential street named Munson Ave.
The GIS-based spatial analysis methodology was as follows: First voting lists were obtained from the city secretary’s office. These lists had the voter’s names, addresses, voting district, and whether or not they voted. It is important to note that the voting districts lined up with major residential subdivisions. Then using the addresses and ARC View’s address matching module the voters (and non-voters) were geocoded. On average 87% of voters were matched with an address the problems were due to either incomplete/incorrect addresses or multiple voters at the same address. Once the voters were geocoded they examined the spatial patterns of actual voter distribution. Followed by conducting a second-order voter distribution analysis using Ripley’s K-function (a common, yet complex formula used to analyze spatial distribution of data).
The First Bond Election was the least controversial and thus was least cared about by the voters. Voter turn-out was extremely low. It received little media coverage, had no local grass roots organization working for or against it, and passed by an overwhelming majority. Voters clearly perceive that it is appropriate to spend tax dollars. Those voters against the bond initiative seem to have been motivated primarily by fear of the potential negative environmental impacts of such municipal service facilities as incinerators and garbage disposal facilities. Contextual/neighborhood effects had a clear impact on voter turn-out, with those precincts slated for the infrastructural improvements having by far the biggest voter turn-out.
Unlike the tax bond issue of 1995, the 1997 referendum on the construction of a city convention center was controversial and received considerable media attention, and grass roots organizations sprang up to attack it. This referendum also confirmed voter clustering, only on a broader scale of between a 1 and 2 mile radius.
The third Issue of re-opening the street Munson Ave. was by far the most controversial issue; it had major media attention and Grass roots organizations on both sides of the issue. The ordinance to re-open Munson Avenue passed by a near 2–1 margin in almost all precincts it had an unusually large voter-turn out. Analysis of the geographic pattern of that turn-out shows it extending in a line south of the TAMU campus. This street provides major north-south relief to traffic. The geographic analysis showed that people in the voting precincts south of TAMU presumably saw the closing of Munson Avenue as restricting their travel behavior and reducing their route alternatives.
This paper shows that GIS-based spatial analysis is immensely helpful in exploring the impacts of voter turn-out on neighborhood effects. On a broader scale this proves that GIS can and should be applied in politics and that politics is deeply rooted in GIS technology and its applications. Enlarge the awareness of political geographers on the potentials of GIS in electoral geographical research. Evidently, the role of GIS has shifted from being simply a data storage/map making tool in the early 1990s to a analytical tool that is used to develop geographic knowledge.

https://lms.southwestern.edu/file.php/3722/Literature/Sui-2002-GIS_Voter_Turnout.pdf

8 comments:

  1. The fusion of GIS and politics is a very attractive concept to me. I think our current electoral system could use all the technological developments and supplements it can get. Mapping out the clustering of voter turnout rates could greatly enhance the campaigning of both local and national candidates. Then again there could be some unintended consequences such as more exploitation of our current political system due to information asymmetry of certain candidates, or the disenfranchisement of certain demographics of people. But nonetheless, the application of GIS in the political world is necessary for the future of both GIS and politics.

    ReplyDelete
    Replies
    1. GIS is heavily used in politics and combined with demographics it is how campaigns target resources. GIS is used today to find not only "battleground" states, but battle group voting precincts in those states. Our vote, especially on the national stage, is really geographically weighted. Our vote counts as much as where we place it. This is why there has been many arguments for moving over to a popular vote model, where every vote is equal and is counted on a national scale, as opposed to a county, precinct scale, which is aggregated to the states level. As long as we do not have a popular vote model, and have our current system, GIS will always play a huge role in directing precious campaign dollars.

      Delete
  2. The First Bond election is a perfect example of "Not in my Backyard", this GIS analysis actual shows how political engagement increased when environmentally questionable infraction is put in. No one want to live close to a smoke stack. What would be interesting is if this trend is true everywhere, does voter turnout always increase? I would suspect that some districts with less engaged voting populations will end up have more of these types of projects built there. This can go in line of research about social justice.

    ReplyDelete
  3. This kind of information definitely seems like it would be very useful for campaigning where there is a correlation between place and the issues being voted on. I wonder if there are any correlations between place and voter turnout and voting results for non-geographically based issues. It would be interesting to see the ideas in this paper applied to experiments with different campaigning techniques used in different parts of a city to see how it affects voter turn out and voting results. Maybe even at Southwestern with different campaign information disseminated to different dorms?

    ReplyDelete
  4. The application of GIS in determining voter turn out seems to have the potential of helping policy makers understand the voters and the local impact of the policy being voted on. By underlining who is voting on an issue you underline who thinks the election will effect them most, and you can then analyze the geographical data to see why those clusters of people think that. By doing this policy makers could then determine why something didn't pass and what they could change in the policy to make it better for the voters, or see why something did pass and use that information for later policies.

    ReplyDelete
  5. I think that GIS and mapping voter data can give some idea as to how people are going to vote, but it isn't completely determinate. If there's some law that says 'Arizona will be effected by ____ ' or something race or gender specific, then it would be appropriate to see whether voter turnout increases or the sway of voting goes one way or the other.

    ReplyDelete
  6. Could this application be used for the opposite reasons, such as targeting those who don't vote? A lot of media has been saying that Hispanics will be deciding who wins the presidential election this year. However many studies show that Hispanics have a lower voting rate than some other cultures. Could candidates use GIS to know where they should probably do more campaigning. They could use this to get votes that they may have otherwise not get, the people who always vote will continue to vote and probably have been set on who they would choose from the beginning.

    ReplyDelete
  7. I wonder if this application of GIS has been used in looking at different socio-economic characteristics of certain areas in the context of voting. With previous knowledge on voting trends, this could be used to target areas with specific characteristics that are parallel (or opposed) to the politican's views, therefore allowing politicans to use certain campaign tactics in certain areas.

    ReplyDelete