Visa Worker Demand and Multitudes of Skills

Why the technology field may be more heavily targeted for "guest workers" than other fields.

5/28/2005

I keep reading about CEO's of large companies complaining about the alleged need for a higher quota of non-immigrant guest workers with technical skills (H-1B and L-1 visas). This is despite a relatively high unemployment rate for computer and engineering professionals since the dot-com meltdown. The accounting profession appears to have a lower unemployment rate and a higher demand (partly due to Sarbanes-Oxley compliance). I even know a few colleagues who switched from computer programming to accounting because accounting offered more opportunities. Yet, I don't hear CEO's clamoring for more accounting visa workers; they seem to focus primarily on technology professionals. The only other profession that seems to have as much visa worker lobbying as technical fields is the nursing profession.

This brings up the question of why CEO's and business lobbyists are focusing on mostly technical professionals over other professions. Although I think part of the reason is "historical habit" and partly because the Indian government has subsidized a surplus of technology graduates, making a tempting supply for US companies; another reason may be due to the laws of probability with regard to myriad sub-skills.

If you look at most computer-related job ads, you will often see a long list of required and desired skills. These long lists have been the target of many jokes among us techies during hard years. These skills tend to be a mixture particular to that company. One cannot realistically target them in advance with school courses and self-study because one cannot predict the combinations that particular companies will need. Yet companies are eager to fit this eclectic list of skills because they don't want to wait for or pay for training. However, the probability of an exact match is statistically very low and being able to select from a larger pool of talent increases the probability of finding a better match.

To illustrate this, let's perform a statistical thought experiment. Suppose we have a pool of 1,000 citizens and 1,000 potential visa workers (currently most in their home country). Let's also assume that on the average they have even qualifications. We can represent their skills with a table resembling this:

Candidate	Status	Skill 1	Skill 2	Skill 3	Skill 4	Etc...
1	Citizen	A	C	B-	C	...
2	Visa Worker	B+	D	B	B-	...
3	Visa Worker	D	C+	D+	A	...
4	Citizen	C-	A	D	C+	...
5	Visa Worker	C	B-	F	B	...
Etc...	...	...	...	...	...	...
2,000	Citizen	A-	C	B	D+	...

Now consider the situation of a company looking for a technical professional with a set of skills particular to their company. They are more likely to find a closer match if the list is longer. Thus, if the size of the pool (our list) is 2,000 instead of 1,000, then a close match is more likely. It does not matter if the average skill level of individuals in the 1,000 person list is the same as the 2,000 person list; having the bigger pool helps merely because it is bigger.

However, having a larger pool can greatly harm the employment of citizens despite having comparable overall skills.

As the graph above illustrates, the larger the pool of candidates, the lower the chance that the best candidate will be a citizen. Keep in mind that the pool of potential visa workers is not just existing visa workers, but the entire world (minus those who cannot qualify for visas). Thus, the chances of a citizen being the best candidate greatly diminish if the employer can comb the entire world for potential candidates.

Pro-visa lobbyists often claim the "problem" is the faulty American education system. I find this claim flimsy. Computer programming does not heavily use subjects from the formal education system anyhow, and yet is highly targeted by visa applications. Any fellow software techie can tell you that employers rarely even care about school grades after one has a few years of experience under their belt anyhow. It is paid experience in the specific tools and computer languages they seem most interested in, and that is the issue being explored here. (Many of us feel that the heavy focus on specific tools rather than general ability may be misplaced, but that is another topic. Those doing the hiring get the final say in selecting what factors to weigh.)

The Gap

One could perhaps argue that the cost of bringing over visa workers would offset minor statistical advantages of a larger pool. However, first, there are often other factors that mitigate that, such as the fact that because most visa workers are young and without family responsibilities, they can and do work long hours. The technical field is notorious for age discrimination, and family responsibilities may be part of the reason.

However, another issue is that the "pool gap" may grow fairly wide as the number of skills involved grows large. The pool of potential programming languages and technical tools, for example, is in the thousands. We have been talking about the "pool" of worker candidates, but another factor is the large pool of potential skills. Few other professions have such a large pool of requested sub-skills (especially rapidly changing ones), and this is perhaps the biggest reason for the demand for visas from companies despite relatively high citizen unemployment in technical fields.

I don't yet have any direct evidence for this, but strongly suspect that the larger the pool of sub-skills involved, the larger the gap between the best match and secondary matches. In other words, if the skills pool is large, then having a larger worker candidate pool makes a bigger difference in match score. It is roughly analogous to purchasing more lottery tickets in order to increase the chances of winning.

I am currently working on a computer simulation to help verify this relationship, but if any of you readers know some shortcut math or prior research, then please let me know about it. (I will post the results here if/when I have them.)

Summary

One of the reasons for the heavier demand for visa workers in technical professions over other professions may be the "multi-sub-skill" factor whereby the probability of a good match increases with a larger pool of candidates, and this phenomenon is magnified by a larger total pool of potential requested skills. However, a larger pool of candidates may create unemployment for citizen technical workers despite having comparable skills. This could be a primary reason why both sides, employers and employees, complain about the visa worker programs in fields that require many sub-skills, such as technology.