Thursday, June 21, 2012

Pace Study's Analysis of Fairview Fire Tax Rate is Flawed

Pace University's Michaelian Institute for Public Policy and Management released its 189 page Fairview Fire District Consolidation and Efficiency Study final report on June 12. This work, known locally as the Pace Study, examines the feasibility of Fairview consolidating with one or more neighboring fire districts. Pace Study Principal Investigator Michael Genito will present this work at a public meeting this evening, according to the Pace Study website.

In spite of the central importance of tax rates to fiscal analysis, the final report devotes only three sentences and one chart to Fairview's past and future tax rates. Unfortunately, these three sentences, which pertain to average yearly tax rate increase and projection to 2017, are incorrect. Also, the chart contains some incorrect data and an incorrect linear approximation. When I presented my analysis to Genito, he readily concurred that all these statements and the chart are flawed.

Flawed Final Report Passage

The flawed information, on page 175 of the final report PDF (labeled page 167), is as follows:
The Fairview Fire District tax rate has increased on average 4.4% each year from 2008 through 2012. A linear regression of the past five years going forward indicates that by 2017 the tax rate would approximate $6.50 per $1,000 taxable assessed valuation. As such, and all things being equal, the median home would expect to see their fire service property tax to rise from $1,321 per year to $1,502 in 2017.

The above chart, copied from the final report, is confusingly labeled “Tax Rate per $1,000 Assessed Value”, but it is clear from context that this data is really tax rate per thousand dollars of market value, otherwise known as true value tax rate. This is the appropriate kind of tax rate for this analysis.

2008 Fairview Fire Tax Rate Is Incorrect

The key flaw is that the 2008 tax rate in the above chart is incorrect. Fairview's effective 2008 tax rate is $5.16, whereas the above chart shows it to be approximately $4.83. The final report's error in Fairview's 2008 tax rate leads to all the other errors in this passage, as will be explained below.

Genito's Blunder

How did Genito come to make this error? He apparently took an unwarranted shortcut. Instead of dividing Fairview's tax levy by Fairview's market value (the correct method, and the definition of true value tax rate), he took the Poughkeepsie portion of Fairview's tax levy and divided it by the Poughkeepsie portion of Fairview's market value. Under ordinary circumstances, such as between 2009 and 2012, Genito's method would give the same — or nearly the same — result as the correct method. Unfortunately, Fairview's circumstances in 2008 were far from ordinary.

Inequitable Apportionment

Long-time followers of my work know that for every year from 2001 to 2008, apportionment of Fairview's fire tax levy between Poughkeepsie and Hyde Park has been inequitable, resulting in different true value tax rates for the Poughkeepsie and Hyde  Park segments, in violation of New York State Real Property Tax Law. In 2008, the Poughkeepsie segment had a true value tax rate of $4.83 — the number on Genito's chart — but the Hyde Park segment had a whopping true value tax rate of $5.96. All these facts were documented in detail four years ago here, and especially here.

Corrected Chart

In order to fairly graph tax rates, the Y-axis should ordinarily begin at zero dollars. Genito's chart begins the Y-axis at $2, presumably to better visualize small changes in tax rate. The following chart, using the corrected 2008 value, takes this decision further, beginning the Y-axis at $5. This way, small changes in tax rate can be seen even better.

Although the final report's chart includes a straight line approximation to the data and an extrapolation to 2017, such analyses are not appropriate to the corrected data. That's because the corrected data simply does not fit a straight line well enough to justify such an approximation. The corrected data cannot meaningfully be used to linearly extrapolate Fairview fire tax rate out even one year — let alone five years. Once again, Genito concurs with this judgement, which is supported by generally accepted criteria for goodness of fit to a straight line. What this means is that there is simply no basis to support the second and third sentences in the final report's passage, which project 2017 values.

Fairview's Tax Rate Has Been Trending Down Until 2012

We know that taxes are always going up, right? Well, not in Fairview. Examination of the corrected chart between 2008 and 2011 shows that Fairview's yearly tax rate change has been downward twice and upward only once. Even the single upward change from 2010 to 2011 leaves Fairview's tax rate lower than it was in 2008. A standard linear approximation to Fairview's 2008—2011 tax rate would show a decreasing tax rate, not an increasing one.

Fairview's Tax Rate Has Been Approximately Constant — Until 2012

Fairview's downward trend in the 2008—2011 time period is actually quite small. It would probably make more sense to approximate Fairview's tax rate during this time period as a constant value. With such an approximation, Fairview's 2008—2011 tax rate is $5.10 plus or minus 1.2 percent for every year in this interval. The 2011 tax rate is equal to this constant value to within 0.2 percent.

Fairview's 2012 Tax Rate Breaks the Pattern

This pattern of constant tax rate is broken in 2012, where the tax rate soars 12 percent from its historical value of $5.10. It is this break from the pattern that makes it infeasible to predict future tax rates. Another way to look at it is that there is no way one could have predicted Fairview's 2012 tax rate by extrapolation from the previous 4 years.

Average Yearly Tax Rate Increase Is Misleading

What about the first sentence in the final report's passage (average tax rate increase of 4.4 percent per year)? This statistic depends crucially on the 2008 value. With the corrected value, the average tax rate increase is only 2.6 percent per year, not 4.4 percent. Thus the passage's first sentence is incorrect.

Of course, even the corrected sentence is of dubious value. Averages can be deceiving. Why mention a formally correct “average increase” when the tax rate actually decreases as often as it increases. A man drowned in a river whose “average” depth was 6 inches. But he was in the 10-foot part. For the average yearly tax rate increase, essentially all of the tax rate increase during the 5-year period occurred in the last year.

Flawed Passage Is Best Removed

According to Genito, the report's inclusion of the above-quoted passage stemmed from a request by Fairview officials (the “Study Committee”) for a projection based on a 5-year history. Now that Genito has accepted my correction, he and I seem to agree that no future projection can be justified by the data. As I see it, the average tax rate increase is misleading as well, and is best omitted. The only part of the flawed passage that could be of positive value is the corrected chart. This chart is certainly useful for understanding Fairview's fiscal situation, but such an understanding appears to be outside the scope of this report.

Wednesday, June 6, 2012

Why Has Fairview's Exempt Percent Increased?

The Fairview Fire District has the highest fire tax rate in Dutchess County, and possibly the highest fire tax rate in New York State. One of the reasons for Fairview's high tax rate is that roughly half of Fairview's market value is exempt from paying fire taxes, but the exempt half still accounts for half the fire and emergency service calls to the Fairview Fire District. Fairview's taxable property owners pay not only their own share of fire tax, but also pay for service to the exempt properties. Fairview residents and property owners have had a longstanding interest in knowing exactly what percent of Fairview is exempt, and how Fairview's exempt percent may be changing over time. Unfortunately, there has been a history of misstatement of Fairview's exempt percent, which I have attempted to correct. See Fairview Fire District's Exempt Percent Is Misstated — Again.

Fairview's Exempt Percent

The following table shows a 5-year history of Fairview's exempt percent, according to my calculations:

Year of Tax BillExempt Percent
* Listed exempt percent for 2008 is after correcting for a blunder by the Town of Poughkeepsie Assessor's Office.
Note that the 2008 “land” tax bill (including the fire, town, county, and other taxes) corresponds to the 2007 assessment roll, and so forth. This table shows that for the years 2008 — 2010, Fairview's exempt percent has been 47.7 plus or minus 0.2 percent. However, beginning in 2011, Fairview's exempt percent has noticeably increased, standing at 51.7 percent for 2012 tax bills. What accounts for this 4 percent increase in Fairview's exempt percent over two years? This post will examine this question.

Exempt Value has Increased While Taxable Value Decreased

Property values have been falling every year in Dutchess County since the 2008 economic meltdown. Fairview's taxable market value fell 11.8 percent between the 2010 and 2012 tax bills. If Fairview's exempt market value had also fallen 11.8 percent during this period, Fairview's exempt percent would have stayed the same as 2010, at 47.9 percent. But Fairview's exempt market value did not fall 11.8 percent — it actually increased by 2.5 percent! In terms of dollars, Fairview's exempt market value for the 2012 tax bill was about $74 million greater than it would have been if Fairview's exempt percent had remained constant. This $74 million caused Fairview's exempt percent to increase from 47.9 to 51.7 in two years. The $74 million arises from two sources:
  1. Four exempt parcels saw dramatic increases in assessed value, for a total of about $41 million. 
  2. Fairview's other exempt parcels fell in value by only about 4 percent on average, rather than by the 11.8 percent decrease for taxable parcels. These exempt parcels are assessed at approximately $33 million more than they would have been if they'd depreciated in proportion to taxable parcels.
Why did four exempt parcels dramatically increase in assessed value?

One might assume that the dramatic increases in four parcels simply reflect major construction projects on these parcels. Surprisingly, this assumption is true only for one of the four parcels. Marist College's parcel at 30 Fulton Street (Parcel number 134689-6162-05-035776-0000) increased in value from $240,000 in 2010 to $17,760,000 in 2012 because student residence halls were constructed on that property between those years.

The other three parcels, detailed in the following table, are part of the water and sewer systems for the City and Town of Poughkeepsie:

Parcel number
Address3431 North RdKittredge Pl173 Kittredge Pl
Land use class822 (water supply)853 (Sewage)853 (Sewage)
OwnerCity of
and Town
City of
Town of
2008 tax bill$4,274,000$111,400*$159,300
2009 tax bill$4,274,000$235,000*$316,000
2010 tax bill$4,274,000Not in roll$316,000
2011 tax billNot in rollNot in roll$300,500
2012 tax bill$12,000,000$10,000,000$5,250,000

The above three parcels have had no significant construction or other actual increase in market value in many years. These parcels have just been improperly assessed for at least 5 years:
  • The 835569 property was erroneously listed as taxable rather than exempt for the 2008 and 2009 tax bills, as indicated by * after its assessed value.
  • This same property was erroneously listed with land use class 340 (Vacant land located in industrial areas) for these same years.
  • Two of the three parcels were erroneously omitted from the assessment roll corresponding to the 2011 tax bill. One was erroneously omitted from the assessment roll for the 2010 tax bill.
  • None of the assessed values for any of these three parcels for any of the 5 years is remotely correct. Even for the 2012 tax bill, the total assessed value is $27,250,000 — only a fraction of the true value of these three properties.
According to my discussion with Town of Poughkeepsie Assessor Kathleen Taber, the 2012 values are only the beginning of an attempt to correct the assessments for these parcels. A realistic correction will not be in place until the 2013 land tax bill, which is based on assessments being finalized this month. The current Parcel Access database, applicable to 2013, shows a tentative total assessed value for these parcels of $125,000,000 — nearly $100 million more than this year's assessment.

Why did exempt parcels decrease in value less than taxable parcels?

While taxable parcels fell in value 11.8 percent in two years, most exempt parcels fell only about 4 percent. As I understand Taber's explanation for this, many exempt properties are difficult to assess because they don't generally appear on the open market. People don't generally buy or sell municipal sewage treatment plants, college academic buildings, or hospital atriums. Changes in the market value of these properties are difficult to gauge because there really isn't a market for these properties. Taber also mentioned that “commercial” properties tend to decrease in value more slowly than residential properties.

It may also be that less attention is given to properly assessing exempt properties simply because the stakes are lower. For taxable parcels, property owners pay real money proportional to the assessment. Taxpayers want assurance that they are paying no more than necessary, while municipal governments receiving taxes want assurance that they are collecting the full amount of money from every taxable parcel. Therefore, tax assessors are under considerable pressure to make assessments of taxable properties that are neither too high nor too low. For exempt properties, these incentives are not present. Inaccurate — apparently even wildly inaccurate — assessments aren't so much noticed.


There isn't one simple answer as to why Fairview's exempt percent has increased in the last two years. According to my analysis, there are three contributors, in order of decreasing importance:
  1. Although Fairview's taxable market value fell by 11.8 percent, Fairview's exempt market value fell by only about 4 percent. The difference means that Fairview's exempt properties were valued $33 million higher than they would have been if they had tracked the taxable decline.
  2. Three municipal water and sewer parcels were grossly under-assessed. The assessor made a correction of $23 million.
  3. Marist College built student residences, increasing the value of one parcel by $18 million.
These three factors contribute 45%, 31%, and 24%, respectively, to Fairview's increase in exempt percent. Thus, all three factors contribute significantly to Fairview's increase.

Pattern of Under-Assessment of Exempt Properties

The careful reader will have noticed two reasons why Fairview's exempt percent may not be as meaningful as one would like. The first is that gross under-assessment of high-value exempt properties is a bigger issue than previously assumed. Two years ago, I found that the St. Frances Hospital complex had been under-assessed by over $100 million. At the time, I assumed this blunder was a one-time event that would be unlikely to be repeated. Now there's a second instance:  Municipal water and sewer parcels have been under-assessed by over $100 million. Most of this under-assessment will not be corrected until Fairview's 2013 tax bill. This pattern will continue: The recent construction of dormitories at Dutchess Community College — worth tens of millions of dollars — will not be reflected in Fairview's 2013 exempt percent. Taber told me she didn't have time to add the DCC dorms to the current assessment roll, the basis for Fairview's 2013 tax. The omission of such major contributors to exempt value results in underestimation of the true exempt percent.

Unequal Depreciation

The second reason why Fairview's exempt percent may not be so meaningful is that market forces apparently do not affect taxable and exempt properties equally. Nearly half (45 percent) of the increase in Fairview's exempt percent in the last two years is due to the fact that the average exempt property lost only one third as much value as the average taxable property did. At least, that's what the assessment rolls say. Do the assessment rolls accurately reflect exempt property values? There is some reason to wonder. If exempt properties have been overvalued in the last few years, Fairview's corresponding exempt percent is artificially high.