Access to Information Orders
Decision Information
BACKGROUND The Municipal Property Assessment Corporation This appeal relates to the manner in which property assessments are calculated in connection with municipal property taxation. Assessments are prepared by the Municipal Property Assessment Corporation (MPAC), which was established by the Municipal Property Assessment Corporation Act, 1997 . MPAC started operating on December 31, 1998, when the government of Ontario transferred responsibility for property assessment to this new organization. Originally named the Ontario Property Assessment Corporation, it became "MPAC" as a result of amendments included in the 2001 Ontario Budget. MPAC is a non-share capital, not-for-profit corporation. Every municipality in Ontario is a member of MPAC, and the organization is governed by a 15-member Board of Directors appointed by the Minister of Finance. The Board includes municipal representatives, property taxpayers and members representing provincial interests. MPAC is funded by its member municipalities. Property Valuation Process MPAC administers a uniform province-wide system, based on current value assessment, and carries out its activities in accordance with the Assessment Act , the Provincial Land Tax Act , and the regulations under these statutes. In documentation submitted in this appeal, MPAC outlines the approach it uses in establishing current values for properties, and how it applies this approach to value in mass appraisal. MPAC explains that it uses advanced statistical techniques and a statistical tool known as "Multiple Regression Analysis" (MRA), and that it estimates unknown data (e.g. market value) from known and available data (e.g. sales prices and property characteristics of sold properties). MPAC indicates that MRA is the tool used to implement the sales comparison approach, and is not the valuation methodology itself. According to MPAC, it uses the sales comparison approach and MRA in a 3-step valuation process: sales investigations and data collection model specification/model calibration model application Sales investigations and data collection In step #1, MPAC analyses and stores data concerning properties and sale information. To do so, MPAC establishes market area and neighbourhood boundaries to be used for analysis and comparison purposes. These are areas referred to as "models", and MPAC explains that there are approximately 165 models in Ontario, 11 of which are in the City of Toronto. The models are geographic areas that are considered to be subject to the same economic influences and are usually, but not always, geographically contiguous. Within these models, locational neighbourhoods are created to capture the influence of location within a given market. MPAC identifies that significant resources are expended by it in defining, identifying, monitoring and reviewing these locational neighbourhoods. Furthermore, their boundaries are not static, and are subject to change based on macro and micro economic fluctuations. Model specification/model calibration Model specification is the formal process of developing a model into a formula or equation. This work is done by MPAC staff, who analyse the factors influencing the local real estate market and determine the property characteristics (independent variables) to test in the particular model. MPAC explains that, in order to specify sound valuation models, an analyst must first conduct data analysis based on a study of property sales in the model area, and then exercise professional judgement in establishing the specification for the model. Once the model has been specified, model calibration takes place. Calibration is the process of developing adjustments, known as coefficients, for the particular model, based on market analysis of the property characteristics that are used in the valuation methodology. This process allocates specific values to the various property characteristics on the basis of the sale prices of sold properties. MPAC stores its sales databases and calibrates its models using the statistical software package SPSS. Once the analysis has been completed and coefficients have been identified, the analyst uses the software to create a syntax file. The syntax file, in turn, creates an output file, which includes the model coefficients and standard statistical information. The syntax file, once created, can be used to re-run the analysis on the current sales database, or to run a new analysis on an adjusted sales database through edits to the syntax file. Model application The model application part of the process involves developing values for all properties within a given market area, by programming the model into MPAC's mainframe computer system, OASYS. All variables and data transformations from each model must be entered into OASYS. Each model is assigned a model number, and the model number is used as the basis of valuation for all properties in the model area. NATURE OF THE APPEAL: MPAC received a request under the Municipal Freedom of Information and Protection of Privacy Act (the Act) for the following information relating to the property assessment of the requester's home: I wish to receive the actual regression equation which is used to calculate residential assessments. Specifically, I would like to see the final equation which was used to determine the assessment on my home…. Specifically I would like the following information: The exact regression equation used to determine the assessed value for my house including its functional form (linear, log-linear, etc). Included in this would be the following: A complete listing of the independent (explanatory) variables used in the property equation. Presumably, these variables are, in part at least, those reported on the property profiles which you use and would include information such as site area, year built, and garage type. Which variables exactly are included, and in what form (continuous variables, dummy variables, etc.)? This request can be addressed by providing me with a copy of the variables codebook. The coefficients attached to each variable in the final equation. These are constant and indicate the impact on the dependent value (assessed value) of a one-unit change in the corresponding independent variable. The requester also identified that he did not want to receive a "disk that contains huge amounts of information". He stated, "[w]hat I would like is just one equation that was used to generate the final assessment value (along with the corresponding codebook)". He went on to state, "I should also be able to input the corresponding information for a neighbour's house and generate the assessed value for that house - and the difference between the two houses should be readily visible." MPAC provided the requester with access to certain information relating to variables and specifics for his own property, and denied access to any information pertaining to the models, including coefficients, variables and equations, on the basis of the following e
Decision Content
BACKGROUND
The Municipal Property Assessment Corporation
This appeal relates to the manner in which property assessments are calculated in connection with municipal property taxation. Assessments are prepared by the Municipal Property Assessment Corporation (MPAC), which was established by the Municipal Property Assessment Corporation Act, 1997. MPAC started operating on December 31, 1998, when the government of Ontario transferred responsibility for property assessment to this new organization. Originally named the Ontario Property Assessment Corporation, it became “MPAC” as a result of amendments included in the 2001 Ontario Budget.
MPAC is a non-share capital, not-for-profit corporation. Every municipality in Ontario is a member of MPAC, and the organization is governed by a 15-member Board of Directors appointed by the Minister of Finance. The Board includes municipal representatives, property taxpayers and members representing provincial interests. MPAC is funded by its member municipalities.
Property Valuation Process
MPAC administers a uniform province-wide system, based on current value assessment, and carries out its activities in accordance with the Assessment Act, the Provincial Land Tax Act, and the regulations under these statutes.
In documentation submitted in this appeal, MPAC outlines the approach it uses in establishing current values for properties, and how it applies this approach to value in mass appraisal. MPAC explains that it uses advanced statistical techniques and a statistical tool known as “Multiple Regression Analysis” (MRA), and that it estimates unknown data (e.g. market value) from known and available data (e.g. sales prices and property characteristics of sold properties). MPAC indicates that MRA is the tool used to implement the sales comparison approach, and is not the valuation methodology itself.
According to MPAC, it uses the sales comparison approach and MRA in a 3-step valuation process:
1. sales investigations and data collection
2. model specification/model calibration
3. model application
Sales investigations and data collection
In step #1, MPAC analyses and stores data concerning properties and sale information. To do so, MPAC establishes market area and neighbourhood boundaries to be used for analysis and comparison purposes. These are areas referred to as “models”, and MPAC explains that there are approximately 165 models in Ontario, 11 of which are in the City of Toronto. The models are geographic areas that are considered to be subject to the same economic influences and are usually, but not always, geographically contiguous.
Within these models, locational neighbourhoods are created to capture the influence of location within a given market. MPAC identifies that significant resources are expended by it in defining, identifying, monitoring and reviewing these locational neighbourhoods. Furthermore, their boundaries are not static, and are subject to change based on macro and micro economic fluctuations.
Model specification/model calibration
Model specification is the formal process of developing a model into a formula or equation. This work is done by MPAC staff, who analyse the factors influencing the local real estate market and determine the property characteristics (independent variables) to test in the particular model. MPAC explains that, in order to specify sound valuation models, an analyst must first conduct data analysis based on a study of property sales in the model area, and then exercise professional judgement in establishing the specification for the model. Once the model has been specified, model calibration takes place. Calibration is the process of developing adjustments, known as coefficients, for the particular model, based on market analysis of the property characteristics that are used in the valuation methodology. This process allocates specific values to the various property characteristics on the basis of the sale prices of sold properties.
MPAC stores its sales databases and calibrates its models using the statistical software package SPSS. Once the analysis has been completed and coefficients have been identified, the analyst uses the software to create a syntax file. The syntax file, in turn, creates an output file, which includes the model coefficients and standard statistical information. The syntax file, once created, can be used to re-run the analysis on the current sales database, or to run a new analysis on an adjusted sales database through edits to the syntax file.
Model application
The model application part of the process involves developing values for all properties within a given market area, by programming the model into MPAC’s mainframe computer system, OASYS. All variables and data transformations from each model must be entered into OASYS. Each model is assigned a model number, and the model number is used as the basis of valuation for all properties in the model area.
NATURE OF THE APPEAL:
MPAC received a request under the Municipal Freedom of Information and Protection of Privacy Act (the Act) for the following information relating to the property assessment of the requester’s home: