Forecasting the Return Home of Non-US Citizens with US PhD's

Project: Research project

Description

Overview
The objectives of this project are to 1) generate a baseline forecast of temporary US visa holders with US Ph.D. degrees who return home, 2) build an explanatory model for the decision to return home by these individuals, and 3) use the results of that model to enhance the predictive validity of individual responses made at the time of graduation about their intention to either stay in the US or return home. To achieve these objectives, access to restricted SESTAT data will be used to first generate estimates of the annual number and rates of returnees for all graduates and then numbers and rates of graduates for the countries generating the most returnees. These estimates will be used to forecast (i.e. extrapolate) 10 years out for each of these series using standard hold-out sample and rolling origin approaches designed to select a best or most accurate forecast in each case. Next drawing on labor market theory, the role of national identities and theories of research productivity, an empirical model will be developed and validated for the decision to return. This model will also consider the roll of home country investments in scientific infrastructure over time. The final step of the research focuses on using the decision model to identify systemic biases associated with a graduates initial estimate of their own likelihood of returning home using data from the Survey of Earned Doctorates (SED). This will be accomplished by modeling individual differences in initial articulation of intent to return home and their actual decisions.

Intellectual Merits of Proposed Activities
This project will enhance our understanding of the subject groups decision process and parses out the role of economic, nationalistic and scientific influences on these individuals. Our results should identify if and how changing status of economic and political activity abroad influences individual decisions to return. The project will further the mission of NCSES by providing current estimates of stay and return rates replacing the current approach. Further the work will contribute to our understanding of how early scientific training and the quality of training also influence the decision to return home. Finally it will provide a tool for enhancing sample selection of individuals from SED data for the SDR and ISDR.

Broader Impacts of Proposed Activities
A major focus of this work is to provide policy relevant information and tools to affect individual decisions in order to mitigate the loss of highly trained and productive scientists and engineers that enhance economic innovation and productivity. The first outcome will provide some evidence as to whether or not the return phenomena is stable, growing or in decline over time. This information provides an assessment of the nature and extent of the phenomena and determines if it is one that suggests the need for policy intervention. The second outcome provides a mechanism for linking policy relevant variables (e.g. nature, amount and conditions of financial support for doctoral education, quality of training, role of foreign governments) to individual likelihoods of return after receiving a US degree. Finally the third outcome can help enhance data quality for regularly collected information from the SED by quantification of bias. Thus these adjusted data may provide more accurate reassessment of individual estimates of their likelihood of return and a better understanding of trends. It is also useful to note that this project will support one full time doctoral student who will become proficient using SESTAT data.
StatusFinished
Effective start/end date9/15/1512/31/16

Funding

  • National Science Foundation (NSF): $102,988.00

Fingerprint

Doctorate
Economics
Financial support
Government
Policy intervention
Graduation
National identity
Nature
Sample selection
Research productivity
Quality of education
Predictive validity
Quantification
Articulation
Labour market
Group decision
Economic activity
Political activity
Empirical model
Data quality