FAQs
Question 1: How do I access the code used to generate the data set?
The full code base and its documentation for the paper are available at this GitHub repo
Question 2: Can we provide you with subsets of the matched data without proof of purchase from L2?
Unfortunately the answer is no. The L2 data is a commercial product and we are not contractually allowed to share the data to individuals or institutions that do not have the proper licenses.
Question 3: How do I tell if my institution has a license to the L2 data?
Given the numerous applications of the data, an increasing number of schools have licensed the L2 data. To determine if your institution has a license, you should start by checking with your school library. You can also contact L2 directly.
Question 4: If I obtain a license from L2 or my institution already has a license and I would like you to share the regression-ready data, what is the process?
Before we can share the full data file with you, we need to first confirm with L2 that you or a research institution with whom you are affiliated has purchased the required data license. Please fill out the data request form. Upon submission of the form, we will begin the process of confirming your access to the data with L2 as soon as possible. Requests are typically processed in 7-14 days from our receipt of the submission form. Once we verify that you have a valid license, we will share the file with you in csv format.
Question 5: What if my institution does not have a license to the L2 data and I do not have the resources in my research budget to purchase a license individually?
The L2 license is institution-wide. Our first suggestion would be to see if a number of faculty could pool resources to secure the license. Note that this data will not only be of interest to management scholars but the underlying data has applications across departments within business schools and in other social science fields. That said, if you are interested in securing a license to only use the data we matched in our paper, we encourage you to reach out to L2 directly and see if an arrangement can be reached.
Question 6: Have you matched the L2 data to corporate boards?
We have only matched the L2 data to executives listed in Execucomp. However, to the extent that our matched executives are also on corporate boards (e.g., for example, they are reflected in the database Boardex), you could infer board preferences by using an Execucomp-Boardex crosswalk or some form of name matching. We have plans, however, to match the L2 data to corporate boards, which should increase coverage and include board members who are not also on top management teams as reflected in Execucomp. We will update this webpage when that matching is complete.
Question 7: Some of the variables in the data appear to be non-binary, what do they mean?
The vast majority of variables in L2 are binary which reflect the presence of an interest within the household. However, some variables, such as book buying and magazine purchases, are ordinal and reflect the number of purchases made within the household. Researchers are free to collapse these variables into binary indicators or to exploit the ordinal nature of the variables, depending on their specific research questions. An enhanced csv file which provides more information on the consumer variables in our matched dataset is available here.
Question 8: Is the L2 data a time-varying panel?
The consumer data in L2 is not a time-varying panel. The data we matched to executives is from 2018-2019. Thus, an assumption when using the data would be that general preferences and interests are relatively stable over time, although researchers are able to select which sample windows to use given their specific research questions. We have updated data from L2, as of 2022, and we plan to match this data to executives in the near future, the results of which will also be made available to the research community.