Author:
Dan Wang – Columbia Business School

Interviewers:
Pete Aceves – University of Chicago
Cassandra Aceves – University of Michigan

Article link: http://asq.sagepub.com/content/early/2014/09/18/0001839214551943

Question 1. In the theoretical discussion you focus on the transfer of knowledge from the returnee to the recipients, who are the returnee’s home-country coworkers. You state that returnees are able to better transfer knowledge gained while working in a host-country if they are highly embedded in their new organizations (within their home country) because they are better able to identify opportunities for knowledge transfer. You also emphasize the role of interaction, stating that “greater engagement deepens one’s connection to a given environment, which in turn grants greater access to the resources embedded within the environment.” All of this might lead one to think that operationalizing embeddedness at the level of the home-country-firm would make sense (i.e. how embedded the returnee is within his/her new firm after returning from working abroad). This is how you measured host-country emeddedness, by looking at the embeddedness of workers within the firms in the host-country. Yet you operationalized home-country embeddedness by examining the frequency of a returnee’s interactions with contacts at home (e.g. family, friends, and colleagues) during the time they were abroad in the host-country instead of examining their connections with coworkers at their new workplace (after returning from working abroad). Can you provide some detail into the thought process behind this operationalization?

This is an excellent question. The journey to this operationalization is a testament to the value of understanding one’s research setting. In the course of collecting survey data on skilled returnees, I also conducted interviews with 160 of my respondents. The interviews informed every step of the research design process. On this particular issue, I realized that returnees developed a broker identity by simultaneously engaging, building, or maintaining connections at home and abroad. Indeed, when Stark and Vedres (AJS 2010) develop their concept of structural folds, they refer to individuals in the Hungarian political and economic elite who were about to contemporaneously and deeply engage with two or more different social groups. The idea is similar in the case of returnees – returnees only serve as good cross-border brokers insofar as they maintain two sets of ties.

Question 2. The unique dataset featured in your paper (where you surveyed over 4,000 former J-1 Visa Holders who worked in the U.S. and then returned to their home-countries) must have required significant effort to collect. The process of securing access and implementing the data collection can be quite time intensive. A few related questions:
A) How did you arrive at the decision that you would collect your own data and how did you establish a connection with IntEx?
B) Did you face any major challenges during this data collection process?
C) What would you say are the advantages and disadvantages of collecting your own data in this way?

A) Finding the data for this project was the biggest hurdle. Initially, I knew I wanted to study the role of returnees in stimulating global flows of knowledge, but no systematic dataset lent itself to such an analysis. I considered using the ISI Web of Science data to track international academic careers or cross-country patent data to study inventors, but even beyond several obvious data cleaning issues, any findings from these datasets would be accompanied by highly restrictive scope conditions. To address some of these concerns, I knew that my data had to be heterogeneous. The J1 Trainee and Intern Visa came up in the course of my initial research, and upon digging deeper, I learned that various sponsorship organizations around the country were responsible for issuing them for skilled foreign nationals to work in companies like Google, Facebook, JP Morgan, and countless other U.S.-based companies. The State department website lists a variety of J1 Visa sponsorship organizations, and IntEx was an organization that I was familiar with. Long story short – while an undergraduate at Columbia, I studied abroad in Germany, and IntEx offered an exchange internship program to Germany, which I almost applied for. I reached out, and after a couple of months of slowly earning meetings with higher and higher levels of the organization, I managed to convince the CEO that conducting a survey of their J1 Visa grantees would be beneficial for academic research and their operations as a cultural exchange organization.

B) There were too many to list. I will emphasize one, however. As the sole investigator in this project (I had great administrative assistance from IntEx in managing survey distribution), I was also solely responsible for every question a respondent would have about the survey. The online survey itself ran for about six weeks. Because the survey was so extensive, for those six weeks, I would spend at least two hours each morning answering emails from respondents or potential respondents. I was encouraged by many respondents who were genuinely interested in the research, but there were plenty of survey recipients who also considered any email survey an invasion of privacy.

C) The biggest advantage is control. Because you control every aspect of the research design and data collection, you become the sole expert on this setting. You know everything about your data from the first to the last row. This greatly facilitates analysis. There will always be another person who knows Compustat or the NBER Patent dataset better than you, but if you collect your own data, you will be the absolute expert. For young scholars, I think this is especially important. The biggest disadvantage is uncertainty. I feel fortunate that I was able to make this data collection work – and I have been able to replicate this survey data collection twice since (see below). However, this is a high reward, but high risk strategy. This form of data collection does not just involve collection, transformation, or aggregation from archival sources – it is better described as the creation of data. My biggest fear during the process was that only 1-2% of survey recipients would respond to the survey. This, of course, would have been disastrous, and has happened to many a graduate student before!

Question 3. This article is one of the chapters of your dissertation. Can you shed some light on A) what sparked this dissertation idea, and B) the iteration process between finalizing a dissertation and crafting a submission to a journal?

A) I wrote this dissertation because a minor disagreement turned into a major disagreement. Originally, I wanted to write something about high-growth ventures in China, so after my second year in graduate school, I decided to set up some interviews with technology entrepreneurs in Beijing and Shanghai. Before the interviews, I read work by Saxenian and others who documented the growth of countries like China, India, and Israel that had benefited from the return of high skilled immigrants. The story I heard from my interviews with those who happened to be returnees contradicted many of the accounts I read. In fact, I got the sense that returnees faced unique challenges in trying to bring back skills and resources from abroad. Other than Elena Obukhova’s work (which influenced me greatly in conceiving of some of the ideas in my dissertation), I found little research trying to understand the barriers associated with returnee re-entry. Soon, I realized that many of the ‘positive’ accounts of returnees suffered from success bias and did not serve as a comprehensive, rigorous narrative of a generalizable returnee experience.

B) This was a challenging process because the paper in ASQ was the first chapter of my dissertation, which set the stage for the following 4 chapters. The paper itself is really an amalgamation of the first two chapters. The first initially contained primarily analysis of survey data while the second drew inferences from my interview data. Featuring both data sources into one paper, rather than emphasizing one over the other, ultimately came from a reviewer’s suggestion.

Question 4. One of the opportunities for future research you highlighted is an analysis of knowledge transfer outcomes when employees return from non-U.S. host-countries .Combining this insight with one of your key findings—that the presence of other returnees in the home-country workplace diminishes the positive effects of host-country embeddedness on knowledge transfer—do you anticipate that a heterogeneous presence of multiple returnees from U.S. and non-U.S. host-countries would create different effects?

This is interesting to think about. I think it is important to first distinguish between home country companies that attract returnees from many different countries and those that attract returnees from only one country. I would anticipate the “queen bee” effect in the home country company that only has U.S.-based returnees, for example, because the social comparison is much more uniform and straightforward. However, home country companies that attract returnees from multiple host countries are likely well suited for integrating diverse knowledge to begin with. One does not need to examine returnees to study this question. One way to study this would simply be to analyze the country-of-origin distribution of employees in companies in the U.S. – for example, one could match H1-B visa data (which are by law, publicly available) to publicly listed firm data to determine the proportion of employees in a given firm from abroad, and the distribution of their countries of origin.

Question 5. What question did we miss? Please ask yourself a good question, and answer it.

Are there any follow-ups to this project? Yes! Since I finished the first survey, IntEx, which was the largest J1 Trainee and Intern Visa sponsorship organization in the U.S., merged with the second largest J1 Trainee and Intern Visa sponsorship organization in the U.S. This allowed me to conduct another survey with this other company, which produced a dataset of 5,000 returnees from 101 different countries. In addition, beginning in January 2014, IntEx and I have implemented a longitudinal survey of every individual that applies for J1 Visa through IntEx. We survey them once before they come to the U.S., once in the middle of their work experience in the U.S., and once after they finish their work experience and return to their home countries. In addition, we also have linked surveys from their supervisors in the U.S. concerning different aspects of their performance including their cultural adjustment to the U.S. Thus far, there are 1,400 individuals who have completed this entire ‘set’ of surveys and counting. With these data, I have developed new projects on returnee entrepreneurship, future returnee migration patterns, and returnee home country readjustment. Please email me if you have any questions about this. Thank you!