Related Documents

Variables used in the statistical calibration of the Development Probability Index for Johnson County
(Power Point, Dec. 2006,
3.2 MB)

Logistic Regression Analysis Output used to calibrate DPI for Johnson County
(Excel, Dec. 2006)

A GIS-Based Approach to Forecasting Land Use Change
(PDF, 3.1 MB)

- - - - - - - - - - - - - - - -

Document Archive

 

Development Probability Index

When we aggregate the amount of growth that individual local governments include in their plans, the total exceeds the overall amount of growth the Kansas City area is likely to experience by 2030. This is to be expected, since local plans are for ultimate build-out, and are not constrained to a specific horizon year. In addition, MARC’s plans require forecasts for 2010 and 2020 as well as 2030. 

Paint the Town uses a Development Probability Index to identify the likelihood a particular polygon will actually develop within the next decade. Given that probability, Paint the Town metaphorically “rolls the dice” to see if the polygon actually develops.

In order to create the Development Probability Index, logistic regression was used to determine which of several criteria related to land use and development had a relationship to the location of development in the 1990s, and the strength of those relationships.  MARC staff collected data on the location of development in the region from 1990 to 1999, plus 20 criteria related to development location selection for 1990 and 2000. The criteria include the location of employment centers, the availability of transportation infrastructure, income, the age of housing units, availability of sewers, poverty and race.     

The Development Probability Index is designed to provide an initial estimate of which polygons will develop within each decade.  However, these probabilities can be overridden by members of the Technical Forecast Committee if they know something the Index does not. In particular, where planners know about new developments likely to occur in specific locations, those areas are assigned very high index values (essentially 100 percent), so that Paint the Town necessarily includes them in the development prediction.