Embedded Files
Do certain firms petition for the H1B Visa much more than other companies do?
The H1B petition data shows which petitions were certified, withdrawn, and denied; to clarify, the petition process happens after a candidate is chosen from the infamous lottery and before he or she receives the actual visa. Since there are thousands of firms, I narrowed the data to the top 20 petitioners. The graph below visualizes the percent share within the top 20 petitioners. By implication, the concentration of the petitions would be much more extreme if this framework were applied to the data of all firms.
The top three firms are indeed Indian out-sourcing companies that are often mentioned in the press: Infosys, TCS, and Wipro.
There does seem to be some truth behind allegations of firms gaming the system.
Do employers at such firms underpay their employees (enough to drive Americans out of jobs)?
I was able to find salaries of San Francisco city employees as a sample for American salaries.
The city started differentiating full-time and part-time employees in 2014, so I used that year's data of salaries for full-time employees.
For an apples-to-apples comparison, I narrowed down Infosys, TCS, and Wipro H1B certified petitioners' salaries to those in San Francisco and in 2014.
The means of the three outsourcing companies were similar: $81715, $82616, and $77292 respectively.
The mean of San Francisco city employees was $103505.
There was more variance in SF salaries because employees had much larger range of jobs.
Although this also supports the claim that these firms underpay employees on average, the difference in means may be significant to some and not to others.
How important is the H1B Visa to the business model of such firms (as measured by stock performance)?
Google Trend: Percent change in log of the relative popularity of searching "H1B" to its peak popularity in google searches. The percent change is from the prior week's value.
Stock Price: Percent change of Infosys' stock price from the prior week's value.
Theory: Last year, when oil prices hit $30/barrel, my friends would google "oil" to buy or sell airline companies because jet fuel costs were almost 50% of their expenses and thus an integral part of their business. I applied to same logic to the stocks of the three companies above. If such companies truly relied heavily on H1B visas, investors would price in changes in H1B visa policies.
Limitations: Sophisticated investors probably don't google "H1B." I also looked at the data for the terms, "H1B Trump," on Google Trends, but the differences between the two queries were minimal. Also, an increase in Google Trend could both be a positive or negative change in H1B. Lastly, expectations could already be priced in so there would be even less need for investors to google "H1B".
Between July 3 and July 16, 2016, there was constant press about US Senators clamping down on H1B visas. Both Wipro and Infosys stocks dropped.
From Jan 29 to Feb 4, 2017, a bill was introduced to clamp down the H1B system again during the beginning of the week. However, towards the end of the week, people realized such possibilities were overblown, reflecting the tipping point of Google Trends. The stock prices thus initially dropped then rebounded, almost simultaneously with the rise and fall of Google Trends.
From Jan 29 to Feb 4, 2017, TCS stock prices displayed the same behavior.
From July 3 to July 16, 2016, TCS stock prices rose unlike those of Infosys and Wipro. This is probably due to the fact that Infosys and Wipro are enlisted on the New York Stock Exchange, whereas TCS is enlisted on the National Stock Exchange of India. Therefore, different macro-economic factors may have played a role in this difference.
Conclusion
By looking at the answers to the three questions above, we can conclude that there are certain abusers of the system. Specifically, Infosys, TCS, and Wipro petition much more than others; they seem to marginally underpay an "average Joe" in San Francisco; and investors believe that their business models do, in fact, rely on the H1B visa system.
My data product could help Congress with policy making. They should target these specific companies, instead of trying to paint the problem with a broad brush.
Google Sites
Report abuse