ECON2206 Introductory Econometrics House Valuations Stage: Term 1 Assessment Answer
UNSW Business School/ Economics ECON2206
Major project: House Valuations Stage Term 1, 2020 Response to request for further data
The Head of the Housing Group at CHS Consulting has responded to the request for further data. The following has been extracted from their reply:
“Housing Group thanks Management Services for assisting them to refine our project. The discussions with your staff have helped us considerably in deciding what needs to be done given the data we have available for analysis. While several of the approaches you have suggested would no doubt provide useful information, we are constrained by data availability. As such it is possible to supply some, but unfortunately not all, of the extra data that was requested by Management Services.
We have decided to undertake a pilot study using recent data for Melbourne, Victoria. As you correctly noted, there is likely to be differences across real estate markets and across different types of dwellings. But we feel that an investigation using the Melbourne data, concentrating on free standing houses is likely to provide a template for a methodology that will be transferable across markets and dwelling types. The data represent a sample of sales of houses in the Melbourne Metropolitan area during 2016. The data includes many house characteristics and importantly a selection of the largest real estate agents. By supplying these largest agents, the data includes most sales and facilitates reliable analysis by individual agents if required.
The discussion between Housing Group and Management Services recognized that valuations are not provided to the public for all houses that are put up for sale and we have complied with your request to provide data that includes all sales irrespective of whether the valuation information was available or not.
Importantly, we confirm that there are three central questions to be addressed using these data:
- Is there any substantive difference between houses for which valuations are publicly provided and those where valuations are not available? Why is this important for subsequent analysis of the underquoting?
- Can you provide a general econometric approach that investigates the accuracy of real estate price valuations? In initial discussions with real estate agents they emphasized the difficulty in predicting the final sales price. This together with the interest amongst policy makers to be prescriptive about providing price guides in terms of a price range we would like you to concentrate your analysis on a measure of accuracy that is defined by whether or not the final price falls within a 10% range of the valuation.
- Is there evidence of differences in the accuracy of real estate price valuations in the data that has been provided? Does this translate into evidence of underquoting by real estate agents?
The Housing Group looks forward to receiving your report.”
The following steps will assist you in your approach to the project.
- Review the project and what is required. Everything that follows should be viewed relative to the key aims of the analysis.
Initial data analysis
- Briefly investigate the key features of the data. (Use descriptive statistics, graphs, etc.)
- The first of the three questions asked by Housing Group requires some basic analysis that forms part of the initial analysis before the other two more substantive questions are addressed.
- Start thinking about how the data at hand can address the second and third questions.
- Develop a possible modelling approach that is relevant for the problem at hand and the data that has been made available that can address question 2.
- Develop an associated model specification. You need to be able to justify the choice and to anticipate potential problems you might encounter in the modelling process.
- Now undertake the analysis making any modifications necessary if you should encounter any problems and use the results to address question 3.
- It is natural to link different parts of the project to corresponding coverage in lectures and material contained there will provide useful references. However, remember you are writing a professional report for management, not providing a sequence of answers to assignment problems.
- Think about how best to report the results. Again, refer to the key questions.
- The report should include a brief Executive Summary. This needs to be non-technical so that Head of Housing Group can assess what you’ve done and what are your basic conclusions. The Chief Economist of CHS is familiar with the more technical components of the report and will advise management on whether the conclusions discussed in the Executive Summary are based on sound econometric analysis and hence credible.
- While the project grade will be based primarily on the substance of your econometric work, the presentation of the material will also be considered. Reports should be typed. We are simulating a professional work experience.
As there would be in an actual work environment, there is no ONE RIGHT way to approach such reports and hence there is no rigid template for what your project should look like nor is there only ONE way to approach the analysis.
These data are realistic but fictious having been constructed for the sole purpose of use in teaching. Data is provided in a Stata file stored on the course webpage in the Assessment Section under Project Stage 2: valuations.dta
inner west east north stheast size carspace bathrms bedrms built cbd AA AB AC AD AE v_yes vbase pbase id Obs: 1678
A description file, valuations.des provides descriptions of each of these variables. Also, in the Project Stage 2 section is a template .do file named valtemp.do . This includes instructions on how to read the data and to produce a group-specific data file. Most of the data will be identical for everyone in the class but several key variables will be generated in valtemp.do . Extra variables price, value and accuracy will be generated by you as a function of pbase and vbase and will be specific to the group. This will mean that key regression results will vary across groups even if they perform identical analyses involving these three variables.
While this provides examiners a plagiarism check this is not the primary reason for providing the data in this form. Instead, it is an opportunity for students across groups to discuss common modelling issues without necessarily coming to the same conclusion.
Naturally you should consult your tutor or lecturer at any other time should you have any other questions regarding the project. This is best done through the online Q&A sessions or by posting a Discussion Topic.
This guided task is a key component in helping students gain some appreciation of what is involved in undertaking a basic econometric analysis.
You are only expected to use the techniques developed in lectures up until the end of week 7.
As a guide your project should be no longer than 2000 words, excluding tables and graphs. Total length including all discussion, tables and graphs should not exceed 12 pages. This is a guide and there will be no penalty for minor discrepancies. You must attach a copy of the FIRST page of your Stata log-file (do-file and relevant output) as an appendix. This first page should include the edited version of the commands in valtemp.do. where you generated your groups’ data. This page will not be included in the word or page count but failure to include this material will result in a zero grade.
Students may work in groups to conduct their analysis and complete the project. Students can choose their own groups and so there is no requirement for group members to belong to the same tutorial. Groups may be up to a maximum of 4 students and students may decide to work alone.
Attach the project assessment sheet provided in Project Stage 2 section as your first page. Check that the names and student IDs for your group members have been completed.
The project will be allocated 25 marks and will constitute 25% of the total grade for the course. The marking criteria are as follows:
ANALYSIS (14 marks)
Motivation for proposed econometric approach & models (3 marks) Characterizing key features of the data (4 marks)
Production and evaluation of econometric results (7 marks)
DISCUSSION (6 marks)
Justification of main conclusions (3 marks) Executive summary (3 marks)
MECHANICS (5 marks)
Presentation (5 marks)