ECON6380/7380/8380 Replication Paper Assignment

Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit

ECON6380/7380/8380 Replication Paper Assignment

1 Introduction

The two assignments in this course are tighly linked to each other. In Assignment One you do a review of a paper published in a reputable economics or econometrics journal. In Assignment Two, you will need to replicate the empirical results of that paper. Please email each assignment to me at [email protected] by the due date and time.

Many journals now require authors to upload their data and even estimation code to the journal's online archive. Journal of Applied Econometrics also maintains a replication section where replicated results are published. A list of journals is suggested below, although you are not restricted to this list alone.

Your task is to look through scholarly journals for empirical work that is of interest to you. For the purposes of this course, it is preferred that you choose a paper with a macroeconomic application. However, exceptions will be made on a case by case basis, if necessary. Please see the instructor if you are not sure whether the empirical application in the paper you wish to replicate is appropriate.

Read carefully through the papers and identify the key empirical questions pursued by the authors, the methodology used, and the results / conclusions obtained. Once you have chosen a particular paper to replicate, think carefully about how you will proceed in replicating the results. Search the data archives of the journal to obtain the dataset used in the paper. Many authors will also include estimation code either in the archive or on their personal website(s). If no estimation code is provided, try to identify what statistical / econometric software package was used in generating the results.

You may will be expected to replicate some important part (not all) of the results exactly as in the paper — i.e., using the same methodology, dataset, code or software, etc. However, you will also need to verify the robustness of the claimed results by attempting to replicate them with a different (but theoretically appropriate) methodology, a different data set (e.g. observations from a different country or add other variables, etc) or extended (e.g. longer observation period) dataset, or alternative software packages.

● Use the headings in this document for each assignment.

● If you only extend the data set, you may pass but will not get a top mark.

● Clarity of writing style is important. If you do not write clearly and concisely it will be very difficult for you to demonstrate that you understand what you are doing.

● Use and label your figures and tables fully and clearly.

● Report all results necessary to support your discussion.

● Clearly identify and compare in tables or figures the original results and your efforts at replicating those results.

2 Suggested Journals

The following journals frequently publish macroeconomics-related research and are known to maintain data / replication archives; they also require authors to upload data and / or code related to all published empirical work.

● Journal of Applied Econometrics

● Journal of Money, Credit and Banking

● American Economic Review

● American Economic Journal: Macroeconomics

● The Review of Economics and Statistics

3 Assignments

Assessment of the replication exercise will be based on two assignments over the course of the semester.

Nota Bene: Before choosing your paper and commencing Assignment 1, be sure you will be able to complete Assignment 2 for that paper. That is, be sure you have access to the correct data and code, and that you can run that code using that data.

3.1 Assignment 1: Due 17 September 2021

Once you have selected a paper to replicate, please complete Assignment 1 by answering the following five questions:

1. What is the paper you intend to replicate (please provide a full, proper reference)?

(a) This is just a reference and so should not be longer than 2 or 3 lines.

2. What are the key objectives of the paper?

(a) This could take between one or two paragraphs and maybe one page.

3. What is the empirical methodology used in the paper?

(a) This is probably going to be the longest part of the assignment. This could take up to between 2 and 3 pages. You are not likely to do this in under one page.

4. What are the key empirical results / conclusions obtained in the paper?

(a) This part may take up to a page, but probably not much more.

5. Please provide a brief description of how you plan to replicate the empirical results.

(a) Here you will outline where you will get the data and code. If you are doing something different, then discuss what you do different, e.g., extra or different data, different methodology, or different software.

The whole assignment one should not be over five pages in length. Write concisely but write clearly and address the most important points.

The purpose of this assignment is to ensure that you are on track with the general replication exercise. Please be as concise as possible in answering the questions above. While you need to be concise, you also need to write well. You will be assessed upon your demonstrated un-derstanding, and understanding is best demonstrated using a good writing style. Responses must be submitted via email to: [email protected]

3.2 Assignment 2: Due 29 October 2021

Please compose a writeup of the replication exercise as a formal paper. The structure of the paper should be the same as in a standard scholarly article. There is no limit on the number of pages; however, a typical replication paper does not exceed five pages, and you should try to remain within this limit.

You may refer to your Assignment 1 answers in part in completing the Assignment 2, but work from Assignment 1 will not contribute to your marks for Assignment 2. Please complete Assignment 2 by answering the following five questions:

1. As in Assignment 1, state clearly the paper you intend to replicate (please provide a full, proper reference)? This should be the first thing you report.

(a) This is just a reference and so should not be longer than 2 or 3 lines.

2. Report your efforts in replicating the important original results. Do not attempt to replicate all of the results, just one or two of the most important tables or sets of results. Did you succeed in replicating the results of the original paper? If not, how did your results differ? Why do you think your results differed? Discuss and clearly demonstrate the similarities and differences. You may provide conjectures as to why the differences exist.

(a) This could take between one or two paragraphs and maybe one page.

3. What extension or difference did you employ in your assignment compared to the original paper and why did you do this? The reason why should be about what the extension will show about robustness (or otherwise) of the results.

(a) This part may take up between one half to a page, but not more.

4. What are the key empirical results / conclusions you obtained? Report in a clear manner the estimation results and give clear interpretations/discussion of these.

(a) This is probably going to be the longest part of the assignment. This could take up to between 2 and 3 pages. You are not likely to do this in under one page.

5. Please provide a brief analysis of the implications of the results from your assignment; pay particular attention to how your results differed from the original paper.

(a) This could take between one or two paragraphs and maybe one page.

6. Prepare a short video in which you present your assignment. This should cover each of the parts 1-5 above.

● This should be no longer than 10 minutes, and possibly only 5 minutes long.

● You may record this on Zoom or on your phone.

● If the file is too large to email, please save on a Dropbox folder and share that folder with me.

● Make good use of figures and tables. The presentation should be clear, concise and not simply be a 'talking head' presentation. Do not simply present the written assignment.

The whole assignment one should not be over five pages in length. Write concisely but write clearly and address the most important points.

For a good outline of what a replication paper should do see (if the links do not work directly, you can cut and paste them into your browser):

Pesaran, H. (2003). Introducing a Replication Section. Journal of Applied Econometrics, 18(1), 111-111. Retrieved from http://www.jstor.org/stable/30035191

A good example of a replication paper done by a Ph.D. student can be found at:

Chan, J. C. (2005), Replication of the results in 'learning about heterogeneity in returns to schooling'. J. Appl. Econ., 20: 439-443. doi:https://doi-org.ezproxy.library.uq.edu.au/10.1002/jae.847

https://doi-org.ezproxy.library.uq.edu.au/10.1002/jae.847

A few good sources of information on how to conduct a replication and to write a replication paper are listed here:

https://scholar.princeton.edu/sites/default/files/bstewart/files/soc504handout.pdf

https://projects.iq.harvard.edu/files/gov2001/files/papertips.pdf

http://ftp.iza.org/dp2760.pdf

This website is designed to help you find a paper to replicate:

http://replication.uni-goettingen.de/wiki/index.php/Main_Page

The paper may be submitted in either hard copy or via email to: [email protected]. If you choose to submit electronically, please ensure the document is either in .pdf (strongly preferred), .doc or .docx format.