Here are a few observations that we did not appreciate before starting this study.

• The bordering communities usually do not include the communities that are most like Arlington in any one characteristic.
  
• There is a big difference between Arlington and the bordering communities in most of the 18 characteristics.
  
• Arlington tends to rank in the middle of the bordering communities as measured by the 18 characteristics (refer to charts on following pages).
  
• No community is similar (among the 6 most like Arlington) in more than a few of the 18 characteristics (refer to charts on following pages).
  
• Although much data is available on web sites sponsored by public interest groups and both State and federal governments, these data may not be directly comparable  or up to date. For example, Arlington includes the cost of crossing guards in its school budget while other communities may place this expense in the police budget.
  
• It is difficult to weigh the relative importance of the varied characteristics such as median family income, population, education budget, number of housing units, though we believe each is important.
  

The existence of a standard set of peer communities is doubtful.
A reader should carefully evaluate the choice of communities when comparisons are made.

How similar are the communities in other ways? (If, for example, a proposal requires an expenditure, do the communities to which Arlington is being compared have similar wealth?)
     If the comparison is to the average (or median) of a group of communities, does the whole group share similar values for relevant characteristics? (If, for example, a public transportation system improvement is proposed, doe s the group chosen for comparison have area and population density characteristics like Arlington?)
Where did the data presented come from? How reliable is this the source of data?

Despite out conclusion, it is likely that officials, consultants and others will still use the community comparison approach. We suggest they consider the following common sense observations.
Massachusetts communities outside the Greater Boston area are much different from Arlington in size, population density, and wealth.

• The communities chosen should be both better and worse than Arlington in the characteristics of interest in the study.
  
• Communities should be chosen so that relevant characteristics other than the characteristic of interest are similar (for example population density when a study recommends expansion of public property).
  
• If bordering communities are used, they should all be used.
  
• Source and age of date should be provided.
  

Massachusetts Department of Revenue

Massachusetts Department of Housing and
Community Development

Massachusetts Department of Education

Metropolitan Area Planning Council

Municipal Benchmarking LLC

NEXT PAGE >