Cluster Analysis

I've been working on an interesting project the past few weeks, clustering of Utah's cities and towns. Often our analysis here at ULCT requires grouping or clustering of cities. But usually this grouping is dictated solely by population size or regional location--very surface and misleading variables. In an attempt to find a more accurate grouping, ULCT has attempted clustered cities together based on demographic, financial, and economic characteristics. The variables used to group cities are:

• 2005 population
• Percent population change 2000-2005
• Household median income (2000 Census data)
• 2003 Primary residential land value (& per capita)
• 2003 Commercial and industrial land value (& per capita)
• 2003 Property tax revenue (& per capita)
• 2003 Sales tax revenue (& per capita)

Our results led to 11 clusters of cities, with Salt Lake City as the one outlier. The eleven types of clusters are: major cities, commercial centers, high growth communities, residential transitioning, high income residential, urban edge, resort communities, natural resource/mining based, old established, traditional agricultural, and small towns.

If you are interested in where your city grouped or would like to see the analysis write-up please review the report: 2007 ULCT Cluster Analysis I'm interested in hearing your feedback on this project. I think we will find it a great tool enabling better policy analysis. Please contact me with any questions.