27 January 2019

How Trump Won (yes won) the Shutdown and What We Can Conclude About Immigration, Income and Crime

The general consensus is that Trump lost the government shutdown. I think he won it. Before I explain why, let's look at some data.

There is nothing like data to undermine certainty.
Donald Trump and Ann Coulter believe that more immigrants means more crime and higher unemployment and / or lower wages. Let's take a look.

First, let's look at a smattering of cities with a population between 200,000 and 300,000. 

Median household income varies greatly, from about $34k a year in Buffalo, NY to $96k in Irvine, CA. Irvine's population is about 40% foreign-born, 10X Buffalo's 4%. Irvine's income is nearly 3X as high.

The correlation between these two variables - income and immigration -  is not perfect but is positive through most of the cities. Immigration and incomes rise and fall together.

What about violent crime? Surely it will rise as the percentage of immigrants goes up, no?

Well, in the above table we can again look at the two cities with the highest and lowest percentage of immigrants to see how crime and immigration are correlated. In Buffalo, violent crime is 179% higher than the national average. That is nearly 3X higher. By contrast, in Irvine violent crime is 86% lower than the national average. (It could only be 100% lower for the simple reason that once violent crime drops to zero it cannot go any lower. 86% lower than the national average is kind of amazing.) We can, again, look at a graph to see a line that is the best fit through all those points.

It is obvious that factors other than immigration change crime rates but as the percentage of immigrants in a community rises, crime falls. 

What about the ten biggest cities in America, you ask. Immigration might be good for mid-size cities but what about cities of millions? (And as it turns out, only the country's ten biggest cities have populations of more than a million.) Well, I have a table for that as well.
Of America's ten biggest cities, Philadelphia has the lowest income and San Jose has the highest. And as it turns out, Philadelphia also has the lowest percentage of immigrants and San Jose has the highest. Immigrants make up only 13% of Philadelphia's population and 39% of San Jose's. Median household income in San Jose is nearly $100k and in Philadelphia is just over $40k. San Jose has 3X the immigrants and double the income.

Above is the graph plotting the relationship between these two variables for the cities over a million. 

Finally, we take a look at the relationship between the percentage of foreign born and violent crime rate in America's biggest cities. Chicago is the most violent of America's biggest cities and 21% of its population was born outside the US. San Jose is the least violent (its violent crime runs 6% lower than the national average) and has 39% immigrants.  The graph looks like this.

Now there are a few arguments you could make when faced with this data. One, you could say that immigrants move into more affluent or peaceful cities but don't help to create affluence or safety. Perhaps the best cities would be even better if not for the percentage of immigrants who move there. The data moves together but immigration doesn't cause higher incomes or lower crime, you say. Perhaps. The fact that the median home price in San Jose is over one million dollars and in Philadelphia is only $158k suggests that it is harder - not easier - to move into these safer, more prosperous areas. 

Or you could argue that immigration has a fairly weak correlation to income and crime, even if it is in the right direction for pro-immigration arguments. The R-squared measure is a simple measure of how well a line fits through the data; at best (median income and foreign-born % in cities of ~250,000) these move together about 40% and at worst (the relationship between violent crime and immigration in America's ten biggest cities) about 24%. So you might say, "Well sure, it seems positive but obviously other factors are a bigger determinant than immigration." And you are right. Education, infrastructure, research and development investments, culture, and social connections are all factors that matter. Immigration is just one dimension of what makes a city great. But the data nonetheless suggest that it IS one dimension of what makes a city great.

Those are valid - but fairly weak - arguments that you could make to discount the relationship between immigration and incomes or crime.

What is not valid to conclude from this data? Higher rates of immigration lower household income or raises crime. That simply does not fit the data. Given the data you could (sort of) challenge the claim that immigration makes a city better but you could not argue that it makes cities worse.

What does this mean? It means that Congress should ignore Trump's demands that they take immigration more seriously. Why? Because immigration is - at the least - a non-issue and - at most - is actually a huge positive that we should encourage rather than discourage. And in spite of that, Trump has forced House and Senate members to treat immigration as if it is an important issue to address. (They have three weeks to "resolve" the issue before another shutdown could hit.) It simply is not. And this is an argument that I've made recently here. Trump has won the shutdown because he has forced Congress to take a non-issue seriously. He has won because he has managed to change the focus of DC onto what he imagines is real, like getting your parents to lose sleep in order to fight the monster under your bed. It is such a waste of leadership potential to solve imaginary problems rather than real ones. (And more generally, a waste of leadership potential to fix old problems rather than create something new. Every successful company puts more money into new product development than it does product repair.)

In the minds of Ann Coulter and Donald Trump, you could predict unemployment rates based on immigration rates. Immigrants steal jobs, they tell us. So, if one city of two million had no immigrants its unemployment rate would be zero and if another city of two million had a million immigrants, its unemployment rate would be 50%. And of course this is an inane way to think about an urban economy, almost as if you thought that brown bodies and white bodies were affected differently by gravity. When a person buys gas or groceries, the market hasn't a clue whether they were born within a block of that place or half a world away. 

There are any number of issues that congress should consider if they are intent on raising income, lowering crime and making life better. Immigration is not one of them. If anything, the data suggests that Congress should do what it can to increase immigration, not decrease it.

We're suffering from the worst recorded case ever of an old man talking back to his TV. Terrifyingly, making his narcissism seem justified rather than delusional, his TV then talks back to him. Trump is on a closed-circuit loop with Fox news. Facts have little influence on his thinking. He gets his talking points from Fox and then they report on what he has talked about. Like Hendrix's guitar, the feedback just increases the volume and the distortion as Trump talks to FOX (Frightened Old Xenophobes) and FOX talks back to him and Trump's story escalates from a campaign to a presidency to a Monty-Pythonesque tragedy.

My two cents? Congress should ignore his insistence that they treat immigration as a real problem and instead either insist on studies as prelude to policy or even celebrate immigration as a positive. It's time to de-escalate the feedback with facts before we are all made as crazy as Trump or waste anymore time chasing his hallucinations.

Quick note: this data is for foreign born. It makes no distinction between legal and illegal immigration; the two move together.

No comments: