Hebron Dollars & Sense

On January 31, 2009, the town of Winchester will have a sixth budget referendum.  That vote will very likely end the Fiscal Year 2008-09 (FY09) budget adoption process for Connecticut towns.  Attached is an Excel spreadsheet that includes details and summaries of the property tax and spending changes for FY09.  Below are some statistics for that budget season and a few observations based on the data from multiple budget seasons.

Referendum Town Statistics for FY09

There was a record high of 82 towns that had at least one binding town budget referendum and another 3 towns that had a non-binding town budget referendum.  49.6% of budgets were rejected at referendum compared to 52.4% over the prior two-year period.  The number of citizen budget votes per referendum town was 1.72 compared to 1.99 over the prior two year-period.  71 of the 82 referendum towns completed their budget adoption process with a referendum vote, a lower percentage than in prior years.   Sixteen referendum towns did not have approved FY09 budgets by 6/30/2008, the end of the prior fiscal year (18 of 169 towns when considering both town and Regional School District (RSD) budgets).

The median effective tax rate (ETR) increase for referendum towns was 2.75% (it was 2.86% for all CT towns).  Additional revenue from the ETR increase and newly taxable property generated a median tax levy increase of 4.09%.  The tax levy and other revenue supported a median spending increase of 3.62% for referendum towns.  That spending increase includes the town’s share of a RSD budget, if applicable. Median net taxable grand list growth excluding real estate revaluation appreciation was a meager 1.29%.

Regional School District Referendum Statistics for FY09

All seventeen RSDs, which encompass 47 towns, used a referendum to approve their budgets.  The median RSD spending increase declined to 3.99% for FY09 from 4.48% the prior year.  It was more difficult to get RSD budgets approved at referendum this year as 41.4% of budgets were rejected by voters compared to 26.1% in each of the last two years. 16 of the 82 referendum towns had separate referendum votes on their portion of the RSD budget. 29 of the 82 referendum towns had separate referendum votes on their local education budget.

The percentage of YES votes at referendum for the local education portion of the town budget tends to be smaller than the percentage of YES votes for the municipal portion of the town budget.  The median difference is 3.2% less YES votes for the local education budget than for the municipal budget.  The separate votes within the town carry separate spending increases, which tend to be higher for the local education portion of the budget.

Tax Increase Differences of Referendum and Non-referendum Towns

For FY09, the median ETR increase for referendum towns was about one-half percent lower than for non-referendum towns.  Over the last six years, the median ETR increase for referendum towns has been about one-quarter percent lower on average than for non-referendum towns.  When excluding revaluations, the median ETR increase of referendum and non-referendum towns has not been materially different over the prior 6 years.  Referendum towns have been more sensitive than non-referendum towns to the tax shifting of recent revaluations toward homes.  The result is revaluation ETR increases tend to be lower in referendum towns than non-referendum towns.

Revaluation Impacts

For FY09, the total ETR increase for referendum towns was 2.92% (the total can be thought of as a weighted average where the weights are the prior year’s taxes).  25 of the 82 referendum towns either had a revaluation or were phasing in a prior revaluation. As a result of revaluation, not all property types had the same total ETR increase. Due to tax shifting in the revaluation towns, the total home ETR increase for referendum towns was 4.82%.  The tax shifting toward homes and away from motor vehicles netted a 3.66% total citizen ETR increase.

In referendum towns that fully implemented revaluation over the last 3 years, the median additional tax increase for homes due solely to revaluation was 5.6%.  The additional median tax increase was 1.9% when the tax shift toward homes is combined with the tax shift away from motor vehicles.

The revaluation tax shift experienced by homes since FY05 has caused more tax unrest than usual.  To lower the tax burden and ease some of that unrest, revaluation referendum towns often end up with a spending increase lower than what would normally have resulted if there wasn’t a revaluation.  In spite of high home tax increases, a lower than usual spending increase requires a lower than usual tax levy increase and hence a lower than usual ETR increase.  In recent years, referendum towns that fully implementing their revaluation in one year had a median spending increase about 1.1% lower than non-revaluation referendum towns. This resulted in a 1.4% lower median ETR increase on average over the last 5 years (median ETR increase is taken of towns within the year and average is taken of median differences o ver the 5 years).

For FY10, 46 towns are slated to have a revaluation and another 9 towns will be phasing in a prior revaluation.  Although the tax shifting toward homes won’t be as great next year as it has most recently, it will still occur and the basic numeric relationships cited above will still be relevant but differences will be reduced. In referendum towns, the median additional tax increase for homes due solely to revaluation was 4.3% in FY09, but 6.5% and 6.8% in the two fiscal years before that.

Influence of Tax Increase and Spending Increase on Referendum Voting

In general, people are more influenced by the home tax increase than the spending increase when casting their votes in a budget referendum.  Over the last three years, 63.0% of budgets at referendum in towns impacted by revaluation have failed.  Towns not impacted by revaluation experienced a lower referendum rejection rate of 46.6% and yet typically carried a higher spending increase than revaluation towns.

The general assertion above was further supported by examining the correlation of the percentage of YES votes in a budget referendum to both the home tax increase and the spending increase.  This was done for budget referendums over the last three years to generate the Correlation Coefficient statistic.  The Correlation Coefficient ranges from -1.0 to 1.0.  Two variables are perfectly correlated if the coefficient is 1.0 and perfectly oppositely correlated if the coefficient is -1.0.  When the coefficient is 0.0, the variables have no correlation (a random relationship).  As spending increases and home tax increases are decreased, the percentage of YES votes at referendum increases.  Here are the Correlation Coefficients between the percentage of YES votes and:

Spending Increases: -.20

Home Tax Increase: -.50

While there are certainly other factors, such as economic conditions, that influence the passage of a budget at referendum, home tax increases influence voting results more than spending increases.

With higher than usual property tax increases very likely over the next two years, it might be time to pause and appreciate that median ETR increases were lower than 3% the past two years.  Attached is a Word document with some early speculation about the median tax increases for FY10.  The attached Excel spreadsheet is strictly concerned with the past and also includes a section with ETR increases by town and year for the prior decade.

Bill Generous
Windsor , CT