UBC Undergraduate Research

Rapid Transit : Housing Affordability Black, Tom; Doshi, Jenil; Lin, Alvin 2018-12

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
18861-Black_T_et_al_ECON_492_Rapid_trasit_housing_benefits_final_report.pdf [ 312.4kB ]
18861-Black_T_et_al_ECON_492_Rapid_trasit_housing_benefits_final_presentation.pdf [ 6.83MB ]
Metadata
JSON: 18861-1.0387007.json
JSON-LD: 18861-1.0387007-ld.json
RDF/XML (Pretty): 18861-1.0387007-rdf.xml
RDF/JSON: 18861-1.0387007-rdf.json
Turtle: 18861-1.0387007-turtle.txt
N-Triples: 18861-1.0387007-rdf-ntriples.txt
Original Record: 18861-1.0387007-source.json
Full Text
18861-1.0387007-fulltext.txt
Citation
18861-1.0387007.ris

Full Text

UBC Social Ecological Economic Development Studies (SEEDS) Sustainability Program Student Research Report Rapid Transit: Housing Affordability Tom Black, Jenil Doshi, Alvin Lin  University of British Columbia ECON 492E Themes: Transportation, FinanceDecember 2018Disclaimer: “UBC SEEDS Sustainability Program provides students with the opportunity to share the findings of their studies, as well as their opinions, conclusions and recommendations with the UBC community. The reader should bear in mind that this is a student research project/report and is not an official document of UBC. Furthermore, readers should bear in mind that these reports may not reflect the current status of activities at UBC. We urge you to contact the research persons mentioned in a report or the SEEDS Sustainability Program representative about the current status of the subject matter of a project/report”. ECON 492E – Rapid Transit: Housing affordability Assignment 4 Tom Black, Jenil Doshi, Alvin Lin  To analyze the impact of the Canada Line on housing prices in Greater Vancouver, we will use data collected by the British Columbia Assessment Authority (“BC Assessment”), one of the primary sources for real property assessments and information in BC. The data will be obtained through an existing use agreement between BC Assessment and the UBC Centre for Urban Economics and Real Estate. Under the use agreement, UBC faculty and students have access to roll data from 2014 to 2018 in BC’s Lower Mainland, as well as real property transaction data from 2005 to 2018.  British Columbia Assessment Authority BC Assessment is a British Columbian Crown Corporation with a mandate to “establish and maintain uniform real property assessments”, as outlined in the Assessment Act1. Whereas BC Assessment’s primary function is to enable reliable real property taxation, this function can only be served by maintaining an extensive database on all properties in BC. Beyond informing tax jurisdictions throughout the province, this data is often used for research purposes; select datasets are also made available online for public access.  1 More information about BC Assessment is available on their website at: https://info.bcassessment.ca/about-us/about-BC-Assessment  2 For most properties, the roll number is a unique number which identifies a single property. However, there are special cases in which one property has multiple roll numbers or in which one roll number refers to multiple properties.  BC Assessment provides residential property assessments as of July 1 each year, based on market value. To determine the value of a property, appraisers analyze physical features of each property (e.g. lot size, building size, age, location, renovations), past transactions, and market trends. All properties are assessed to their “highest and best use”, which estimates a value based on the property’s potential uses. To determine the highest and best use, appraisers seek to understand the legal permissibility (i.e. permitting), physical possibility (i.e. geographic constraints), financial feasibility (i.e. ability for developers to finance a potential use scenario), and maximum profitability of each property. According to BC Assessment, this methodology provides an accurate estimate for the market value for properties.  1 More information about BC Assessment is available on their website at: https://info.bcassessment.ca/about-us/about-BC-Assessment     Data  Through the use agreement between BC Assessment and the UBC Centre of Urban Economics and Real Estate, we have access to raw roll data from 2014 to 2018 and raw transaction data from 2005 to 2018 in the Lower Mainland. The roll data are the annual assessments which BC Assessment provides, as described above, which includes information on the land/building size, land/building value, building features, etc. for individual properties throughout the Lower Mainland. Whereas the transaction data does not contain information on the physical features of the property, it provides the sale price and assessment values for every transaction. All observations are identified by a roll number which is a unique identifier for each property2.  Methodology  Our primary dataset will be the raw transaction data from 2005 to 2018. We select this dataset because the available roll data does not cover property information from 2009, the year in which the Canada Line was implemented. To uncover our key parameters of interest, the data will be analyzed through a regression analysis using a difference-in-differences model, the baseline of which is outlined below.   𝑙𝑛𝐻𝑃𝑖,𝑁,𝑡 =  𝛼 +  𝛽1𝑇𝑟𝑎𝑛𝑠𝑖𝑡𝑖,𝑁  × 𝑃𝑂𝑆𝑇𝑡 × 𝐶𝐿𝐷𝑖 +  𝛽2𝑇𝑟𝑎𝑛𝑠𝑖𝑡𝑖,𝑁  ×  𝑃𝑂𝑆𝑇𝑡 + 𝛽3𝑇𝑟𝑎𝑛𝑠𝑖𝑡𝑖,𝑁+  𝛿𝑡 + 𝛿𝑁 + 𝜀𝑖,𝑁,𝑡  HPi,N,t refers to the total value of a piece of property, i, in a particular neighborhood in Greater Vancouver (as represented by the subscript N) and a particular year (as represented by the subscript t). We take the natural logarithm of this term to allow us to interpret our βs as a percentage change. Transiti,N is a variable which captures the distance of a property to a rapid transit station (or where a station would be, in the case of Canada Line locations pre-construction). The data for this variable can be obtained by geocoding addresses from the BC Assessment datasets and calculating the Euclidean distance to the nearest station. CLDi is a dummy variable which takes a value of 1 if the Transiti,N variable refers to a distance to the nearest Canada Line station (otherwise, CLDi equals zero). POSTt is a dummy variable which takes a value of 1 after the Canada Line has been opened, allowing us to capture the effect of the policy shock. Since housing prices can change prior to the opening of the Canada Line as a result of speculation, we can conduct a sensitivity analysis with varied point in time for which                                                           2 For most properties, the roll number is a unique number which identifies a single property. However, there are special cases in which one property has multiple roll numbers or in which one roll number refers to multiple properties.   POSTt takes a value of 1 (e.g. when the Canada Line station locations were announced). The terms δN and δt are neighborhood and year fixed effects, respectively. Finally, εi,N,t is the error term. Beta 1 captures the percentage change in housing prices for as a result of opening the Canada Line for those who are closest to the Canada Line. The coefficient on Transiti,N × POSTt (Beta 2) captures the change in housing prices as a result of the opening of a new transit line. The coefficient on Transiti,N (Beta 3) is the percentage change on housing prices caused by proximity to a transit station. To enhance our results, we can consider adding further effects for specific household characteristics.  In order to uncover the effect of the Canada Line on housing prices in different neighborhoods, we can expand on the baseline regression described above. First, the baseline regression assumes that housing prices in each neighborhood have parallel trends. That is, housing prices in each neighborhood are changing at similar rates before and after the Canada Line is implemented. If this assumption does not hold true, we must account for neighborhood-specific time trends by including another term which interacts the neighborhood dummy variables with a linear time trend. Alternatively, we can account for area characteristics such as employment density, crime rates, or income in each neighborhood.  Second, we can augment the baseline regression by including an interaction between our covariate of interest, Transiti,N × POSTt, and each neighborhood. Adding this term to our regression could allow us to estimate a separate causal effect of transit on housing prices in each neighborhood. This is important to capture because the effect of being close to rapid transit on housing prices may be different in the Central Business District (CBD) than in a suburban neighborhood.  By estimating separate effects, we can potentially support the case of extending the Millennium Line to UBC, using the Canada Line as a proxy. Assume a scenario in which housing prices near the CBD far exceed the prices of those that are further removed – a relatively accurate albeit simplified  representation of housing prices in Greater Vancouver. An ideal scenario is if we find that effect of transit on housing prices is negative in neighborhoods close the CBD and positive in neighborhoods which are further removed. Such a result would signify that implementing the Canada Line improved housing affordability by decreasing the disparity between housing prices based on distance from the CBD. Residents close to the CBD now enjoy lower housing prices, residents further away now enjoy lower transportation costs. Furthermore, residents can also choose to move closer or further from the CBD, based on their preference for housing consumption and transportation costs.  UBC Social Ecological Economic Development Studies (SEEDS) Sustainability Program Student Research Report Rapid Transit: Housing Affordability Tom Black, Jenil Doshi, Alvin Lin  University of British Columbia ECON 492E Themes: Transportation, FinanceDecember 2018Disclaimer: “UBC SEEDS Sustainability Program provides students with the opportunity to share the findings of their studies, as well as their opinions, conclusions and recommendations with the UBC community. The reader should bear in mind that this is a student research project/report and is not an official document of UBC. Furthermore, readers should bear in mind that these reports may not reflect the current status of activities at UBC. We urge you to contact the research persons mentioned in a report or the SEEDS Sustainability Program representative about the current status of the subject matter of a project/report”. ECON 492E – Rapid Transit: Housing affordability Assignment 4 Tom Black, Jenil Doshi, Alvin Lin  To analyze the impact of the Canada Line on housing prices in Greater Vancouver, we will use data collected by the British Columbia Assessment Authority (“BC Assessment”), one of the primary sources for real property assessments and information in BC. The data will be obtained through an existing use agreement between BC Assessment and the UBC Centre for Urban Economics and Real Estate. Under the use agreement, UBC faculty and students have access to roll data from 2014 to 2018 in BC’s Lower Mainland, as well as real property transaction data from 2005 to 2018.  British Columbia Assessment Authority BC Assessment is a British Columbian Crown Corporation with a mandate to “establish and maintain uniform real property assessments”, as outlined in the Assessment Act1. Whereas BC Assessment’s primary function is to enable reliable real property taxation, this function can only be served by maintaining an extensive database on all properties in BC. Beyond informing tax jurisdictions throughout the province, this data is often used for research purposes; select datasets are also made available online for public access.  1 More information about BC Assessment is available on their website at: https://info.bcassessment.ca/about-us/about-BC-Assessment  2 For most properties, the roll number is a unique number which identifies a single property. However, there are special cases in which one property has multiple roll numbers or in which one roll number refers to multiple properties.  BC Assessment provides residential property assessments as of July 1 each year, based on market value. To determine the value of a property, appraisers analyze physical features of each property (e.g. lot size, building size, age, location, renovations), past transactions, and market trends. All properties are assessed to their “highest and best use”, which estimates a value based on the property’s potential uses. To determine the highest and best use, appraisers seek to understand the legal permissibility (i.e. permitting), physical possibility (i.e. geographic constraints), financial feasibility (i.e. ability for developers to finance a potential use scenario), and maximum profitability of each property. According to BC Assessment, this methodology provides an accurate estimate for the market value for properties.  1 More information about BC Assessment is available on their website at: https://info.bcassessment.ca/about-us/about-BC-Assessment     Data  Through the use agreement between BC Assessment and the UBC Centre of Urban Economics and Real Estate, we have access to raw roll data from 2014 to 2018 and raw transaction data from 2005 to 2018 in the Lower Mainland. The roll data are the annual assessments which BC Assessment provides, as described above, which includes information on the land/building size, land/building value, building features, etc. for individual properties throughout the Lower Mainland. Whereas the transaction data does not contain information on the physical features of the property, it provides the sale price and assessment values for every transaction. All observations are identified by a roll number which is a unique identifier for each property2.  Methodology  Our primary dataset will be the raw transaction data from 2005 to 2018. We select this dataset because the available roll data does not cover property information from 2009, the year in which the Canada Line was implemented. To uncover our key parameters of interest, the data will be analyzed through a regression analysis using a difference-in-differences model, the baseline of which is outlined below.   𝑙𝑛𝐻𝑃𝑖,𝑁,𝑡 =  𝛼 +  𝛽1𝑇𝑟𝑎𝑛𝑠𝑖𝑡𝑖,𝑁  × 𝑃𝑂𝑆𝑇𝑡 × 𝐶𝐿𝐷𝑖 +  𝛽2𝑇𝑟𝑎𝑛𝑠𝑖𝑡𝑖,𝑁  ×  𝑃𝑂𝑆𝑇𝑡 + 𝛽3𝑇𝑟𝑎𝑛𝑠𝑖𝑡𝑖,𝑁+  𝛿𝑡 + 𝛿𝑁 + 𝜀𝑖,𝑁,𝑡  HPi,N,t refers to the total value of a piece of property, i, in a particular neighborhood in Greater Vancouver (as represented by the subscript N) and a particular year (as represented by the subscript t). We take the natural logarithm of this term to allow us to interpret our βs as a percentage change. Transiti,N is a variable which captures the distance of a property to a rapid transit station (or where a station would be, in the case of Canada Line locations pre-construction). The data for this variable can be obtained by geocoding addresses from the BC Assessment datasets and calculating the Euclidean distance to the nearest station. CLDi is a dummy variable which takes a value of 1 if the Transiti,N variable refers to a distance to the nearest Canada Line station (otherwise, CLDi equals zero). POSTt is a dummy variable which takes a value of 1 after the Canada Line has been opened, allowing us to capture the effect of the policy shock. Since housing prices can change prior to the opening of the Canada Line as a result of speculation, we can conduct a sensitivity analysis with varied point in time for which                                                           2 For most properties, the roll number is a unique number which identifies a single property. However, there are special cases in which one property has multiple roll numbers or in which one roll number refers to multiple properties.   POSTt takes a value of 1 (e.g. when the Canada Line station locations were announced). The terms δN and δt are neighborhood and year fixed effects, respectively. Finally, εi,N,t is the error term. Beta 1 captures the percentage change in housing prices for as a result of opening the Canada Line for those who are closest to the Canada Line. The coefficient on Transiti,N × POSTt (Beta 2) captures the change in housing prices as a result of the opening of a new transit line. The coefficient on Transiti,N (Beta 3) is the percentage change on housing prices caused by proximity to a transit station. To enhance our results, we can consider adding further effects for specific household characteristics.  In order to uncover the effect of the Canada Line on housing prices in different neighborhoods, we can expand on the baseline regression described above. First, the baseline regression assumes that housing prices in each neighborhood have parallel trends. That is, housing prices in each neighborhood are changing at similar rates before and after the Canada Line is implemented. If this assumption does not hold true, we must account for neighborhood-specific time trends by including another term which interacts the neighborhood dummy variables with a linear time trend. Alternatively, we can account for area characteristics such as employment density, crime rates, or income in each neighborhood.  Second, we can augment the baseline regression by including an interaction between our covariate of interest, Transiti,N × POSTt, and each neighborhood. Adding this term to our regression could allow us to estimate a separate causal effect of transit on housing prices in each neighborhood. This is important to capture because the effect of being close to rapid transit on housing prices may be different in the Central Business District (CBD) than in a suburban neighborhood.  By estimating separate effects, we can potentially support the case of extending the Millennium Line to UBC, using the Canada Line as a proxy. Assume a scenario in which housing prices near the CBD far exceed the prices of those that are further removed – a relatively accurate albeit simplified  representation of housing prices in Greater Vancouver. An ideal scenario is if we find that effect of transit on housing prices is negative in neighborhoods close the CBD and positive in neighborhoods which are further removed. Such a result would signify that implementing the Canada Line improved housing affordability by decreasing the disparity between housing prices based on distance from the CBD. Residents close to the CBD now enjoy lower housing prices, residents further away now enjoy lower transportation costs. Furthermore, residents can also choose to move closer or further from the CBD, based on their preference for housing consumption and transportation costs.  Rapid Transit:Housing AffordabilityAlvin Lin, Tom Black, Jenil Doshi“ Policy QuestionShould the Broadway Subway project be extended to UBC? Research QuestionHow does extending the Broadway Subway project to UBC influence regional housing affordability? 2Background3● The Broadway Corridor is one of BC’s more important economic centres○ Second largest employment hub ○ Connects UBC and Vancouver General Hospital● Lack of sufficient transportation○ According to the C.D. Howe Institute, congestion can cost anywhere from $0.5B to $1.4B 4● In Summer 2018, Phase Two of Translink’s 10-Year Vision was approved, confirming the construction of the Broadway Subway ● The Broadway Subway project includes:○ Extension of the Millennium Line from VCC-Clark to Arbutus St.○ Planning for a further extension to UBC (approximately 7 kms)● According to 2016 Conference Board of Canada Report, the net present value of benefits of extending the Millennium line to UBC are estimated to exceed  $4 billion5Evaluating the Benefits of Rapid Transit● Conventional Benefits - Reduced travel costs, reduced operating costs ● Impact on Housing Affordability● Agglomeration Benefits - Productivity gains Literature ReviewBaum-Snow and Kahn (2000):● Added convenience of living near transit station tends to increase the housing prices● The authors argue that potential time savings benefits are reflected in higher housing prices6Glaeser et al. (2008):● Public transit proximity drives housing prices downward as impoverished households are more likely to converge near the transit stations to capitalize on decreased transportation costs● Rich households are willing to incur greater transportation costs to possess more land in order to live in larger houses Our Approach7Data ● British Columbia Assessment Authority ○ Raw Roll Data (from 2014 to 2018) ■ Addresses ■ Floor area ■ Building type○ Raw Transaction Data (from 2005 to 2015) ■ Sale prices ■ Assessment values ● Acquired through a Data Use Agreement with the UBC Centre of Urban Economics and Real Estate8Economic FrameworkIsolating the effect of transit on housing prices▪ Regression analysis of a property’s sale price on its proximity to the nearest transit station ▪ Compare the effect before and after the Canada Line is constructed ▪ Controlling for: ▫ Floor area ▫ Condos vs. Detached homes ▫ Time trends▫ Neighborhood differences10Calculating proximity to the nearest station1. Plot properties and Canada Line stations to calculate distances2. Identify the nearest station113. Use straight-line distances to create five “distance rings” at 200m increments Key variables▪ Sale prices of properties from 2005 - 2015▪ Distance rings at 200m increments ▪ Nearest rapid transit station▫ Helps us segment the Canada Line▪ Floor area ▪ Building type▪ Transaction dates▪ Neighborhood of the property 12Results14Change in sale prices after the Canada Line was constructedSale Prices and Proximity to Transit▪ Prices start relatively high▪ Decrease until 2km ▪ Increase at greater distances▪ Potentially because of the tradeoff between the added convenience and congestion associated with proximity to transit stations15The Canada Line increased average housing prices throughout Metro Vancouver, but the effect varies considerably by region ▪ Binary distance variable▪ Segment areas along the Canada LineTakeaways:▪ Positive effect ▪ No consistent effect of increasing distance ▪ Largest effects in “Vancouver - Residential” areaHow does the effect of proximity to transit on sale prices change after the Canada Line is constructed?16Although average prices increased at the municipal level, certain areas within municipalities saw an opposite effect▪ Five distance rings at 200m incrementsTakeaways:▪ Properties in “CBD” and ”Other” align with narrative▪ Prices increase across the board in the “Residential” area▪ Unclear pattern “Richmond” How does the effect of proximity to transit on sale prices change after the Canada Line is constructed?17The construction of the Canada Line made properties which are closer to public transit more valuable than those further away▪ Price disparity between properties within 1 km and those greater than 1 km away from a stationTakeaways:▪ All positive effects in Ring 1▫ There’s value in living near rapid transit▪ Negative effects after Ring 2 in “CBD” and “Other”How does the average price disparity between properties close to transit and those far away change after the Canada Line was built?Summary of findings▪ The Canada Line increased average sale prices in Metro Vancouver▪ Effect of proximity to transit on housing sale prices is not uniform▫ Properties in “Residential” and “Richmond” affected the most▫ Some properties in “CBD” and “Other” see price decreases ▪ Trade-off between convenience and congestion associated with proximity to transit▫ The extent to which residents value convenience and congestion likely vary by region18ConclusionHow does extending the Broadway Subway project to UBC influence regional housing affordability? 20▪ We preliminarily conclude that the proposed project would have a  ▪ This effect varies by region and proximity21Research Limitations▪ Distance measures and travel methods▪ Net expenditures & access preferences▪ Zoning changes 22Ring Distance Shortcomings23Monthly SavingsPer Trip 1 trip per day 2 trips per dayArbutus >> UBC $2.55 $55.47 $110.93Net Expenditures & Access Preference  This is evaluated using a time value of $11.41/hour  provided by the client● Changes in net expenditures have not been measured● How do transit time savings and housing prices affect utility?24Zoning Regulations● B.C. Government recognizes RRT encourages transit oriented housing● Millenium Line extension may encourage densification● Estimated that zoning Vancouver similar to Langley may lead to 2.3% increase in housing starts (The Fraser Institute)● Land regulation in Vancouver is estimated to have lead to an average $600,000 increase in prices from 2007-2016 (C.D. Howe Institute)25Further Research:● Rent Data Analysis● Network Distance● Consumer Preference studies26Thank you!Questions?AppendixOLS Regression 1 - Binary distances28OLS Regression 1 - Point-estimates29OLS Regression 2 - Distance rings30OLS Regression 2 - Point-estimates31Graveyard (To be deleted)Table of Contents ● Policy and Research Question● Background ● Knowledge Gap ● Literature Review ● Our Approach ● Results and Discussion● Recommendation● Research Limitations & Next Steps33

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.18861.1-0387007/manifest

Comment

Related Items