Skip to 0 minutes and 0 secondsWe know that crime rates rise and fall over time. For example, the rate of violent crime, including homicide, rape, aggravated assault, and robbery in Oakland, California went down by 25% between 1996 and 2014. We also know that there is very big difference in crime rates across places. Even after falling by 25 percent, the rate of violent crimes in Oakland is still more than twice as high as a violent crime rate in Pittsburgh, Pennsylvania. What can explain this variance in crime rates? One obvious explanation is that the differences in crime rates is driven by the differences in demographic, socioeconomic, and criminal justice system characteristics across place and time.
Skip to 0 minutes and 52 secondsIndeed, after many years of research, we know that crime is closely related to a number of factors, such as the size and effectiveness of police, labor market conditions, the quality and quantity of education received, age and racial distribution, gang membership and illegal drug markets and so on. A number of researchers worked very hard to lead to recover the causal effect of these factors on crime. And I think our understanding on the determinants of crime has improved a lot over the past years. However, our understanding is still not perfect. For example, high-crime and low-crime areas usually differ in this demographic and socioeconomic characteristics, but the differences are usually much smaller than the actual difference in crime rates.
Skip to 1 minute and 44 secondsFor example, according to the 2014 U.S. Census, Oakland has population of about 400,000 people. About 20% population are under age 18, about 30% are black, about 80% have high school education or above, the median household income is $54,000, and the poverty rate is around 20%. On the other hand, Pittsburgh has population of about 300,000 people. About 15% of population are under age 18, about 25% are black, about 90% have high school education or above, the median household income is around $40,000, and the poverty rate is at 23%. It’s clear that there are some notable differences between the two cities. Oakland has more population, more young people, more ethnic minority, and have less education than Pittsburgh on average.
Skip to 2 minutes and 47 secondsHowever, none of these differences seems big enough to explain why Oakland has violent crime rates more than twice as high as Pittsburgh. As another example, take the trend of violent crime rates in the United States in the last few decades. This figure shows that the rate of violent crime in the United States has gone up by more than 400% between 1960 and 1990, but then it fell down by about 50% between 1993 and 2010. Again, there have been a number of important changes in demographic and socioeconomic conditions in the United States, but none of them is big enough to explain this drastic increase and decrease of crime in such a short period of time.
Skip to 3 minutes and 35 secondsIn fact, to many of us who study crime, it still remains somewhat of a mystery why crime fell down so much in the 1990s. Professor Steven Levitt, an economist from the University of Chicago, wrote an article back in 2004, titled "Understanding Why Crime Fell in the 1990s". Although we do not have a perfect answer for the crime drop during the 1990s, this article still gives us a great summary of the existing findings on the crime drop during the 90s. And shows us that our current understanding on what causes crime and what can be done to lower crime can actually explain a big chunk of the crime decline during this period of time.
Skip to 4 minutes and 20 secondsWe will take a close look at this article after this video.
Crime disparities across places and over time
Our understanding of what causes a national crime epidemic and decline is still limited.
However, thanks to research evidence accumulated over last few decades, we are now more confident of our understanding of how an individual’s criminal decision is influenced by changes in socioeconomic factors and criminal justice environment.
© Songman Kang, Hanyang University