Skip to 0 minutes and 1 second In recent years, we saw a drastic increase in the volume of empirical research done by economists, driven by the availability of new high-quality crime data and advances in computing power. Covering the advanced empirical methods used by economists is probably out of the scope for this class, but there is still a lot we can learn from the large volume of crime data that’s available out there. Crime is one of the most important policy issues in many countries, and many national and local governments collect data on crimes and criminals, and make the information available to the public and interested researchers.
Skip to 0 minutes and 42 seconds In later weeks, we will see examples so very unique crime data, but for today, let’s talk about two main sources of crime data in the United States. We will focus on the crime data in the U.S. because I have mostly worked on U.S. crime data in the past and know more about it than other places. There are two major data sources for crime in the United States. First, FBI collects data on crime counts from law enforcement agencies across the country, and publishes this crime data annually. This FBI Uniform Crime Reports, or UCR data, is publicly available and widely used by those who do research on crime.
Skip to 1 minute and 27 seconds This is the UCR website where you can learn more about the UCR program, or you can simply search for “FBI UCR” on an online search engine. Here it says they offer four annual publications. And if you click this “Crime in the United States” link, we reach this webpage that has information on the number of crimes, persons arrested, and police employees. You can click these pages and download the information as an excel file.
Skip to 2 minutes and 0 seconds Alternatively, you can build a simple table based on the UCR data from this UCR data online website (http://www.ucrdatatool.gov/). Another major source of crime data is from the Bureau of Justice Statistics. They run a crime victimization survey, called the National Crime Victimization Survey or NCVS. In this survey, they ask about 90,000 households across the United States, twice a year, about whether they were victims of crime, and if so what the details of each crime they experienced. You can find a lot of information about what the survey questions look like, a summary of the survey responses, and research publications based on the NCVS data from the Bureau of Justice Statistics website.
Skip to 2 minutes and 54 seconds The FBI UCR data and the victimization survey data complement each other. FBI data is collected from official reports from police agencies and covers almost the entire country. But it is limited by the fact that it only contains information on crimes that were reported to police. Many crimes are never reported to police, and in many cases, criminals are not caught. Such crimes and criminals will not show up in the UCR data. On the other hand, the victimization survey includes information on crimes that took place even if they were not reported to police. This is a very important advantage of this NCVS data.
Skip to 3 minutes and 41 seconds According to the 2014 NCVS data, only about one third of the victims of rape, about a half of victims of domestic violence, and less than 30 percent of victims of theft report to the police. The UCR data will tell us nothing about these unreported crimes. One limitation of using the victimization survey data is that it is based on a relatively small sample of 90,000 households. The rates of victimization for certain crimes are very low and when the sample is not that big, such survey data can tell us very little about the patterns of such crimes. Also, working with survey data can have complications.
Skip to 4 minutes and 24 seconds For examples, respondence may misunderstand the questions, their memory of the victimization may not be accurate, or they may not answer their questions truthfully, intentionally. In this video, we talked about the two main sources of crime data in the U.S. which tell us about the pattern of crime in the entire country. But many local governments and police agencies also collect data on crimes and criminals and make this information available to the public. You will see examples of such data sets in our next lesson.
Crime data source
These days, many local and national governments publish their data on crimes and criminals online, and finding good crime data is easier than ever.
When reporting crime data, many governments put special emphasis on certain types of “street crimes” that are relatively well-defined and pose great risk to the public safety. For example, in the U.S., the FBI UCR defines seven Part I index crimes (homicide, forcible rape, aggravated assault, robbery, larceny, burglary, and motor vehicle theft), and the crime data you find from the online UCR data tool is based on these seven index crimes. There, homicide, forcible rape, aggravated assault, and robbery are grouped as “violent crimes” and larceny, burglary, and motor vehicle theft are grouped as “property crimes”. (FBI UCR also collets information on Part II index crimes, which include offenses such as simple assault, fraud, gambling, etc.)
While some policy changes can increase or decrease the number of total “crimes” in the absence of actual changes in people’s criminal behavior (e.g., legalization of abortion, decriminalization of marijuana possession), the major crime index based on a small number of relatively well-defined and serious crimes is unlikely to be affected by such policy changes.
In the next step, you will be asked to find some crime statistics publicly available online. Below are a few links that you may find useful for the data exercises.
FBI UCR Page: https://ucr.fbi.gov/crime-in-the-u.s
FBI UCR Data Building Tool: https://www.ucrdatatool.gov
Bureau of Justice Statistics, National Crime Victimization Survey (NCVS) Series: https://www.bjs.gov/index.cfm?ty=pbse&sid=6
© Songman Kang, Hanyang University