Predicting Hotel Tax Assessments in Cook County - By Hans Detlefsen

Introduction

This study examines the relationship between hotel room revenue and the assessment values of hotel properties located in Cook County, Illinois. We examined the relationship between these characteristics to construct a model predicting property assessment value based on hotel room revenue. We then used a regression analysis to determine the statistical significance of the model and its potential applications to the current market. 

The most recent economic recession led to sharp declines in performance at many hotels in Cook County. Because the value of a commercial property, such as a hotel, is typically linked to its income potential, many owners believe their properties have significantly decreased in value throughout this period of economic decline. As a result, the Cook County Assessor’s Office has recently experienced an increase in property assessment appeals by hotel owners. 

Each year, the county reassesses approximately one-third of the commercial properties located in Cook County, including hotels. For this reason, a hotel assessed in 2007 or 2008, at the peak of the last business cycle, may have an assessment value that is much higher than a hotel assessed in 2009 even if the two properties are currently very similar in value and performance. Hotel owners whose properties were assessed in 2007 and 2008 may feel their assessments are unfairly high, and many have filed appeals to challenge the current assessment levels. On the other hand, some properties assessed in 2009 have assessments that may reflect a value that is lower than recent performance statistics would indicate, given the recovery in performance that has taken place through most of 2010. The authors hope that the findings of this article may be a useful tool in making a preliminary evaluation about whether certain hotels are subject to assessments that are inconsistent with their peers and their recent performance statistics.

The complete assessment of properties in Cook County occurs over a three-year period. To examine the variation in hotel property assessment values, we collected data for 92 hotels in Cook County from assessment years 2007-2009. Since the assessor typically examines one district per year, we use a three-year span of data to ensure that each property is reassessed at least once. Property assessments are completed intermittently from March to December of the specified year. Therefore, we used estimated rooms-department revenue data of the complete preceding year in our analysis because this would have been the most recent annual data available to the assessors at the time the assessed value was determined. For example, the room revenue data for hotels assessed in 2009 is from the 2008 calendar year. 

In order to assist in making a preliminary determination of an appeal’s legitimacy, we constructed a model comparing the property assessment value with the hotel’s room revenue. We derived a trend line equation from the model to allow us to make predictions about the appropriate property tax assessment value given a hotel’s room revenue. Further regression analysis confirmed statistical significance between room revenue and property assessment value, suggesting that our model may be a useful tool in examining the variation in assessment levels of hotels in Cook County and other localities.

Data Analysis

Our sample of hotels in Cook County demonstrated the sharp increase in property assessment appeals experienced by the assessor’s office in recent years. In the 2007 assessment year, only 1% of hotels in our sample appealed their assessment value. In 2008, the number of hotels that filed an appeal increased to 12%. Finally, the number of appeals in our sample for the 2009 assessment year more than tripled from the previous year to 45%. The dramatic rise in appeals among hotel owners in Cook County encouraged us to search for a useful tool to predict the appropriate assessment values for a given hotel. This tool may be helpful to officials reviewing tax appeals as a quick, preliminary guide in their evaluation of which appeals may have legitimate grounds for adjustment; moreover, it may be helpful to hotel owners who wish to determine whether or not it would be appropriate to file an appeal.

We initially hypothesized that the net income of a hotel would be a useful predictor of its property assessment value because net income is generally considered the most important factor in determining market value of a hotel. However, data on net income at a large sample of local area hotels is not readily available. Instead, we used rooms-department revenue as a proxy for income because it is typically the largest revenue line item and the most profitable department in most hotels. The HVS database of hotels in Cook County includes room revenue for a majority of the hotel properties located in Cook County. As such, this approach allowed us to have a more accessible income-related variable to analyze in terms of its relationship to assessment value.

The assessment period in our analysis ranged from 2007 to 2009. During this period, the assessor examined each of the three assessment districts in Cook County. The following graph shows the spread of hotel assessment values per room revenue during the three-year period.



The graph depicts a strong positive correlation between hotel room revenue and assessment value. The correlation leads us to hypothesize that there is a meaningful relationship between the two variables, which we can break down through further analysis. As a preliminary step, we sorted the variables according to their assessment year and derived trend lines for the predicted assessment value in a given year. The following graph shows the overall fluctuation in assessment values from 2007-2009.



This graph shows how the relationship between room revenue and assessment value has fluctuated with the market. The fluctuation reflects the changes in the market that occurred during the assessment period. Properties had the highest assessment values-per-dollar-of-room-revenue in 2007, which declined in 2008 and rose slightly in 2009. The fluctuation is this relationship reflects changes in opinion about risk during the same period. For example, when perceptions of market risk were very low in 2007, values-per-revenue were highest. On the other hand, when risk was perceived to be highest, during the trough of the recession in late 2008, values-per-revenue were lowest. These fluctuations appear to be consistent with trends in capitalization rates for hotels during the same period; capitalization rates increased in 2008 and part of 2009, then began to decline again toward the end of 2009 as signs of a stabilizing economy emerged. . Therefore, despite the fact that market perceptions about risk also influence market values, and assessment values, we conclude that room revenue can be used as a good predictor of the assessment value of a hotel.

To further analyze the relationship between room revenue and property assessment values, we ran a regression of assessment value on room revenue. The room revenue variable has a high t-statistic (24.98), which tells us that room revenue is statistically significant in relation to property assessment value. Generally, a t-statistic greater than 2.0 indicates statistical significance at the 95% level. The R₂ equals 0.87, which further strengthens our hypothesis from Figure 1 that a strong relationship exists between rooms revenue and property assessment value. In this case, the R₂ value explains the portion of the variance in assessment level that can be explained by room revenue. We used the following simple, linear regression equation to interpret our results:

Y _i = β₀ + β₁X₁ + u_i 

Y_i = dependent variable (assessment value)
Β₀ = constant term
Β₁ = the slope of the regression
X_i = independent variable (rooms revenue)
u_i = error term
_i = number of observations

The equation of our regression is as follows:

Assessment Value ($Millions) = .79 + .40[Room Revenue ($Millions)]

From this model, we can determine that a 0.4 million dollar increase in room revenue is associated with a 1 million dollar increase in assessment value. As expected, when a hotel’s room revenue increases, so does its assessment value. The following graph is an overlay of the regression line onto the scatter plot from Figure 1:



Since the location of a hotel may have an impact on its assessment value, we controlled for the assessment districts to eliminate any potential omitted variable bias and strengthen our predictions. However, due to the structure of the assessment process, we cannot separate the year variable from the location variable. Thus, the following model necessarily accounts for a hotel’s location as well as its assessment year:

Assessment Value (Millions) = .493 + .40[Room Revenue (Millions)] + .67(North) - .13(South)

Including the location of the hotel in the regression allows us to make more accurate predictions of the property assessment value. We are able to use location in addition to room revenue to explain the variance in assessment levels. We can see from the equation that a hotel in the North district that was assessed in 2007 will have a higher assessment value compared to a hotel in the City district that was assessed in 2009, for a given level of room revenue. On the other hand, a hotel in the South district assessed in 2008 will have a lower assessment value compared to a hotel in the City district assessed in 2009, for a given level of room revenue. Conditional on room revenue, a hotel in the North assessment district is associated with a $0.67 million increase in assessment value relative to a hotel in the City assessment district. On the other hand, a hotel in the South assessment district is associated with a $0.13 million decrease in assessment value relative to a hotel in the City assessment district, conditional on room revenue. These results reflect the fluctuation in hotel assessment value previously illustrated in Figure 2. 

Conclusion

The statistical significance of hotel room revenues in the regression analysis strongly suggests that it can be a useful predictor of property assessment values. This is not a surprising result, as hotels with greater values would be expected to generate greater revenue generally. However, the findings may be useful to owners or assessors trying to develop a preliminary opinion about whether a particular hotel’s property tax assessments are in line with its peers, simply given the hotel’s room revenue, which can be calculated as the product of its occupancy multiplied by its average daily rate. Such a calculation should never be considered a substitute for assessing a hotel’s market value or fair assessment value because many omitted variables need to be considered in the valuation process.

About Hans Detlefsen

HANS DETLEFSEN is Managing Director of HVS Global Hospitality Services in Chicago. He holds a Masters Degree in Public Policy from the Harris School of Public Policy Studies at the University of Chicago, where he received the Harris Fellowship. He graduated magna cum laude from the University of Notre Dame with a Bachelor of Arts in Government and Economics.

www.hvs.com

Reader Comments:

Be the first to leave a comment for this article