Crime rates could take account of the harm caused to victims rather than simply counting offences using a new system, experts have said.
Weighting crimes in relation to their impact would provide a more accurate picture of public safety in the UK, according to criminologists.
They claimed the current approach is a "paper-and-pencil legacy of the 19th century" which presents crime in totals that give equal weight to shoplifting and homicide.
An index devised by researchers from the University of Cambridge aims to incorporate an assessment of the likely impact on the victim of an offence by using sentencing guidelines.
The academics argued their proposals would dramatically improve identification and policing of areas where the most damaging crime takes place.
Professor Lawrence Sherman, director of the Cambridge Institute of Criminology, said: "Not all crimes are created equal. Counting them as if they are fosters distortion of risk and accountability.
"If shoplifting drops while murder triples, crime is reported as 'down' - yet any common sense view of public safety cries out for some adjustment for seriousness.
"Currently, there is no meaningful 'bottom line' indicator of whether public safety is actually improving or declining in any given year or place.
"Measuring by the number of days in prison each crime could attract ensures that police, policy makers and the public are better informed on rates and trends of crime, the risks posed and resources required."
A team including Professor Sherman and Peter Neyroud, a former chief constable, set out how the "Cambridge Harm Index" could work in a paper published in the journal Policing.
It uses a fixed scale based on the number of days of imprisonment an offence would result in at the lowest starting point for an offender with no previous convictions, according to guidelines prepared for courts.
Where penalty guidelines are expressed in community service hours, the index converts them into days and if the starting point is a fine then it calculates how long it would take to earn the sum at minimum adult wage.
The method excludes crime detection which is "proactively generated" through enforcement activity, with the report saying such reports "do not reliably measure harms experienced by the population".
A 10-year comparison between current crime metrics and the proposed method was included in the study.
It showed that, overall, crime counts for a number of selected offences between 2002 and 2012 indicated a fall of 37%.
However, this was said to be an overestimation when compared with a less steep drop of 20% when using the index's measure of "imprisonable" days.
Currently, official crime statistics are based on two main datasets: police recorded figures which cover offences reported to and logged by forces and the Crime Survey for England and Wales (CSEW), in which residents are asked about their experiences of crime in the previous year.
The approach should be added to the current system, rather than replacing it, and the index would require no new funding, the criminologists said.
Last year, the first official estimates of the scale of fraud and cyber crimes were released, sparking claims the overall count could be substantially higher than previously thought.
Overall there were an estimated 6.6 million incidents of crime in the year ending September, according to the CSEW.