Data on crime can come from any of four sources:
Ask the police
Ask the courts
Ask the victim
Ask the offender
Each of these has advantages and disadvantages. Detail follows.
The four sources of crime data Uniform Crime Reports (UCR) (Ask the police) NIBRS - Incident-Based Reporting -- Future FBI reporting Offender-Based Transaction Statistics (OBTS) (Ask the courts) National Criminal Victimization Survey (NCVS) (Ask the victim) Drug Use Forecasting (DUF) (Ask the offender)
Reform orientation of a police department affects the crime rate; the fewer crimes you ignore, the higher the crime rate.
Side issues may affect the crime rate; the more crimes you ignore, the lower the crime rate.
A Georgia police officer: "It burns me up that New York City doesn't report larcenies to the FBI unless they involve $3000 or more; they think complete recording would lower tourism and affect their economy. Meanwhile, we report every theft of $1.35 from a gas station."
Hierarchy rule: if more than one offense is committed in a particular incident, only the most serious offense is recorded in the UCR. For example, assault/rape/murder is recorded only as a murder for UCR purposes. The less severe offenses are ignored.
Participation by local police and sheriff's offices is totally voluntary. Data are thus not complete. Data may thus also be biased.
Many important crimes, and their impact on people are totally ignored by the UCR Part I index.
None of these offenses in Part II is counted as part of the "crime rate".
"Other" is the largest single category on this list (see Table 2.2, p. 55). If the largest category is "other", you haven't defined your categories very well. You don't really know what's in your very largest category.
Will shift to incident-based reporting; for each incident will collect detail on when it occurred, the kind of weapon used (if any), the type and value of property stolen or damaged, the age, sex, and race of both victim and offender, the relationship between the offender and the victim (if any), whether an arrest was made, etc.
National survey is conducted through ISR at the University of Michigan. Includes broad national sample plus intensive samples of largest major metropolitan areas. (For methodology students, this makes it a stratified random sample.)
Any individual respondent is included for seven consecutive six-month waves. (For CJ251 students, this makes it a perpetual panel.)
Questions are phrased so as to describe offenses in common language rather than legal language.
Response rates are consistently high, with over 90% of arrestees approached agreeing to be interviewed. Approximately 80% of those interviewed provide a urine specimen (i.e., approximately 72% of those approached).
Self-report study. Instead of objective data, we're going to ask folks what they've done--whether or not the police have heard about it.
Main NYS runs through IBS at University of Colorado; other smaller samples may be taken elsewhere for other purposes (Boys Town Follow up, for example--unconnected to NYS, but used NYS methodology and some of their questions)
Since respondents may well admit to offenses in the survey that they haven't actually been arrested
Should not even dream of doing this kind of study without being granted a confidentiality certificate by the federal government provides you as a researcher with protection in case you are asked by police to reveal your data in court.
Ethics says you can't reveal your data; certificate, if it has been granted, says you probably won't go to jail for standing on your ethical position. You hope.