Shopping for Politicians: Insights from Market Basket Analysis of Senatoriables

Document Type


Publication Date



At the supermarket, we often buy things that go together—pasta and spaghetti sauce, beer and chips, ice cream and cones, and so on. Certain pairs or sets of items are frequently bought together in the same basket or cart, be it physical or virtual. Using big data analysis, online retailers like Amazon.com exploit these relationships in their recommendation engines, based on items that are “frequently bought together” and those characterized as “customers who bought this item also bought”. Looking for relationships between pairs or sets of items that tend to be purchased together is a data mining technique known as market basket analysis. While many factors drive consumer decisions, this method of analysis reveals common patterns of consumption by aggregating purchase data of millions of customers. Aside from the retail setting, market basket analysis is being used to uncover patterns and associations in events such as voting and elections. For instance, an examination of voting patterns in the House of Representatives of the United States showed how representatives are more likely to vote for or against specific issues based on to their political party (Bagui, Mink, and Cash, 2007). In the same manner, the technique has been utilized to highlight connections among an individual’s age, residence, political views, race, and TV viewing habits using election survey data in the U.S. (MacDougall, 2003)…