Parapundit's Randall Parker has an encouraging post on ICE's stepped up effort at deporting foreign criminals serving sentences in US jails and prisons. But the number of immigrants in need of removal according to Julia Myers doesn't mesh with estimates of the foreign-born prison population being touted by open borders advocates like Linda Chavez. I've looked through some of the relevant data and wasn't able to conclude much, but I might as well offer what I came up with. I suggest you read RP's entry before dealing with what's below.
The average time served for a prison sentence handed out at the state level is just under three years, and it's even longer than that at the federal level. It's unusual for someone held in a state or federal prison to serve less than one year behind bars. So I don't see how sentences of less than one year could be accounting for the huge disparity unless the bulk of the deportation orders are for those who are being held in local jails, where sentences are usually for less than one year.
It doesn't seem to add up. The study by Ruben Rumbaut that Steve Sailer referred to (via Linda Chavez) doesn't give absolute numbers of prisoners broken down by ethnicity and nationality, but it does give percentages. Extrapolating from that, the study estimates that there were fewer than 70,000 foreign-born males between the ages of 18-39 in state or federal prisons in 2000. Males between the ages of 18-39 comprise around 70% of the prison population, so we can probably assume that the total population of incarcerated foreign-born under Rumbaut's methodology comes to 100,000.
ICE is going after criminal immigrants at all levels. Local jails hold one-third of the nation's total incarcerated population, but Rubenstein's figure of 290,050 from the Department of Justice report only applies to state and federal facilities. If we assume that at the local level the demographics are similar to what is seen at the state and federal levels, somewhere in the area of 130,000 of those slated to be deported are being held at the state or federal level. This is nearly twice the total number of foreigners that Rumbaut estimated were being held in all jails across the country. That difference would have had to come in the course of only six years.
Two-thirds of the US foreign-born population is from Latin America. As Rumbaut's report shows, foreign-born Hispanics are more likely to be incarcerated than are European or Asian immigrants. So it'd be pretty safe to assume that of those 130,000, around 100,000 are Hispanic, yielding an estimate that about one-in-three Hispanic prisoners in the US are foreign-born. Yet Rumbaut's total incarceration figure for not only the Hispanic foreign-born but other non-native prisoners and which includes local jails, totals about that amount.
At least one of the following must be true:
1) Rumbaut's study was faulty.
2) Myers' numbers are wrong.
3) There was a significant increase in the size (and rate) of the foreign-born prison population between 2000 and 2006.
4) Local jails hold a greater number of immigrants than do state and federal prisons, and most of those slated for deporation are currently being held at the local level.
Local sentences vary by location, but the average length of a felony sentence to be served in a local jail was just over six months in 2000. Of those convicted of felonies in state courts, 41% of those sentenced to jail time served it in a local jail.
For Rumbaut's estimates and Myers' numbers to compute (ignoring the six year time gap for simplicity), most of the deportation orders--at least 100,000 or so of the 200,000--would need to not only be coming from the local level, but also be for those who have been given sentences of less than one year. Why? Because Rumbaut's numbers are pulled from a specific point in time, when the Public Use Microsample (PUMS) for Census 2000 was performed. If 100,000 or so foreigners had been in jail sometime during the year but were not incarcerated when the survey was carried out, both Rumbaut and Myers would theoretically be correct.
Since there are more felons being held in state prisons than in local jails, the majority of those 100,000 or so being slated to be deported from local jails (and quite likely a majority of all those being slated for deportation, since I'm assuming that ICE is going to get every foreign-born felon held at the state and getting the non-felony number by plugging in what's left over) would have to be serving time for non-felonious crimes (DUI, drug possession, etc).
On its face, this seems highly unlikely. Myers could probably clear up the confusion if she were queried on what percentages of these 200,000 people are currently being held in federal, state, and local facilities, but if I had to put money on it, I'd wager that either Rumbaut's estimate is too low or Myers' is too high. I'm not sure why ICE, which hasn't exactly reveled in the idea of having to step up deportations, would overstate the number of foreign criminals in need of removal. I can imagine why others might, though, but am not able to prove they're wrong, either.
Parenthetically, Rumbaut shows that the longer immigrants have been in the country, the more likely they are to be behind bars (see table 3). The putative rate of incarceration triples for foreigners who've been in the US for more than 15 years as compared to those who've been here for less than five years. This meshes generally with Steve's point that immigrants, who are disproportionately poor themselves, are assimilating to the cultural norms of America's native poor.
But it also stems from the fact that a 30 year-old who arrived stateside yesterday has had a lot less time to get himself thrown in a US jail than have natives who've lived here their entire lives. Yet as soon as he enters the US, he's 'counted' among the non-criminal foreign-born population. And if he has a criminal past at home that isn't known in the US, he's likely to spend less time in jail (and therefore less likely to be counted among the incarcerated at any given time) when he does commit a crime than would a native with a record who committed the same crime.
++Addition++On RP's thread, John S Bolton observes that deportations of foreign criminals, even though they've been inadequate (ICE reports that last year it identified 64,000, most of whom were deported), will also artificially lower the incarceration rate. It's similar to what was discussed above except on the other end--instead of receiving a lighter sentence since the crime was the first one the immigrant had committed (in the US), after serving a relatively light sentence, the foreign criminal is deported, thus keeping him from receiving a relatively harsher sentence for subsequent crimes based on his criminal record in the US.