Are Illegal Immigrants Everywhere? Will it Change Your Vote?

November 01, 2010, 12:42 PM GMT+0

Meg Whitman, California GOP gubernatorial candidate, admitted to hiring an undocumented worker as a housekeeper for nine years. This sort of thing seems to happen every so often. However, you have to wonder why anybody gets upset: by the estimate of most Americans it would be pretty difficult not to hire illegal immigrants. Let me explain…

In 2008, the Pew Hispanic Center, calculated that there might be 11.2 million illegal immigrants in the United States. That is about 4% of the population. The state with the greatest proportion was Arizona, with 7.9%. These numbers, like any Census of a large population, are necessarily inaccurate, but they are the best we’ve got. With the recession, most experts agree that the population of illegal residents has probably declined since 2008.

If you read my last blog posting, you may not be surprised to know that poll respondents have a completely distorted estimate of the number of illegal immigrants in their state. The average response when asked what percent of their state is illegal immigrants was just over 17% - that is, the average American thinks almost 1 in 5 people are illegal immigrants. If that were the case, Meg Whitman might be able to use simple demographics as an excuse. You may also remember that 17% was actually also the average response when asked what percent of their state had an income of over $250,000. One (loose) way to interpret this is that, unless the average American thinks that many illegal immigrants make over $250,000, then the implication is that the average American also thinks that over 1/3 of people in the United State are either illegal immigrants or making over $250,000.

This is quite the distorted view of reality – more than 400% mistaken on the number of illegal immigrants. However, it is not really unexpected. Scholars have repeatedly found that individuals are bad at answering these sort of questions – for example, Cara Wong has shown that although African Americans make up only about 12% of the U.S. population, survey respondents put the number at over 30%.

However, it is not entirely clear why people are bad at estimating this sort of thing. Some evidence points to people simply using the information from their immediate area to guess at the population for the entire country. That doesn’t seem so unreasonable: outside of academics, most people don’t have a need to study Census tables, so they just use whatever information is available to them – in this case it might be personal experience. My survey evidence indicates that this might be what people are doing because the actual (estimate) of illegal immigrants in a person’s state is a pretty highly correlated with their guess – that is, the more illegal immigrants actually residing in their state, the higher the response when asked about how many illegal immigrants resided in their state. Of course, this doesn’t exactly explain why people are so bad at answering this question because, as mentioned previously, no state ever had more than 8% illegal immigrants in recent history.

In addition to living in a state that actually has a lot of illegal immigrants, another thing that makes you likely to believe there is a lot of illegal immigrants in your state is being politically conservative. Statistical analysis that controls for other factors, indicates that people that describe themselves as politically “very conservative”, rather than “moderate” think that an additional 13 percentage points of illegal immigrants live in their state. Of course, the causal direction is difficult to sort out here – do political conservatives believe that they are surrounded by illegal immigrants because they are political conservatives or are they political conservatives because they believe that they are surrounded by illegal immigrants?

Does this distorted view of reality have any electoral consequence? It does not appear so, at least in the short term. Despite reports that illegal immigrants were an electoral concern in places as seemingly unaffected as Nebraska, my statistical analysis indicates that people perceiving a higher presence of illegal immigrants in their state are no more or less likely to vote Republican on Tuesday nor does it seem to affect their opinion of President Obama, once party and other influences are controlled (the very small number of Nebraskans in my poll gave an average answer of about 12% illegal immigrants in their state – the actual number in 2008 was estimated to be less than 3%).

One final note – Americans of all political ideologies show some tendency to, not only over-estimate the number of illegal immigrants, but also to have trouble separating Hispanics generally from illegal immigrants particularly. Let me explain: in addition to asking about the percentage of illegal immigrants in their state, I also asked what percentage of their state is Hispanic. I think it is safe to assume that when most people hear the term “illegal immigrant”, they almost exclusively assume that the illegal immigrant is from Latin America - meaning that when they say 17% of their state is illegal immigrants that they are saying that at least 17% of their state is Hispanic. If I am correct about this, a rough way to estimate the percentage of Hispanics that a respondent thinks are illegal immigrants is simply to compare their answer to the questions about the percentage Hispanic to their answer about the percentage illegal immigrants. Among all white respondents, the average answer to the question about illegal immigrants was almost 2/3rds the size of their answer to the question about Hispanics – meaning, if they thought about it – they are likely indicating that they think that 2/3rds of all Hispanics in their state are illegal immigrants. Among “very conservative” whites the answer to the Hispanic question was, astoundingly, almost 3/4ths the size of the answer about illegal immigrants – implying that 3/4ths of Hispanics are illegal immigrants.