Despite these data, there is, however, often a strong perception that high immigration levels and high crime rates are associated. Indeed, the recent high levels of immigration have coincided with the highest rates of incarceration in modern times. The correctional population—which includes persons in prisons and jails, on probation, and on payroll—was less than 1 million in the United States from 1925 until about 1970. It rose gradually from about 1 million to about 2.5 million in 1980, and more rapidly since then, reaching over 5.5 million in 1995.20
Several good reasons suggest, however, that the temporal association of high immigration and high crime rates is coincidental and not causal. Some, but not all, of the increase in overall crime rates is due to the increase in the number of young people in the population as the baby-boom cohorts entered their teenage and young adult years. Similarly, crime rates have leveled off, and they have actually begun to decline in the mid-1990s, although immigration remains at very high levels. The rise in the imprisoned population is partly due to changes in sentencing policies, resulting in longer prison sentences; most of the increase results from an increase in arrests, principally connected with drug crimes. Finally, using data from crime reports and Current Population Surveys, Butcher and Piehl (1996) concluded that an influx of recent immigrants into a community has no association with local crime rates.
Although U.S. history gives evidence of ambivalence toward immigration of long standing, polling data gathered in the decades since World War II provide more detailed evidence on how the American public regards immigrants and immigration. In recent decades, the overall trend has been toward more opposition to immigration, but with frequent oscillations, as illustrated in Figure 8.3. The fraction of Americans who say they think immigration should be decreased from the current level has risen from less than half of the population in polls taken up until the mid-1970s, to roughly two-thirds of the population in more recent polls.
To try to understand this shift, we analyzed a set of polling data from 1995 that included some questions on immigration. Our analysis focused on respondents' answers to the question, ''Should immigration be kept at its present level, increased, or decreased?" The multivariate analysis included variables to capture
There is no simple link between crime rates and the size of the correctional population. There are many filters between a crime and imprisonment, including the reporting of the crime, the apprehension of the criminal, the sentencing of the criminal to prison, and the length of incarceration. At each stage, variations can occur.