Researchers from the University of Cambridge examined crime statistics both before and after the introduction of the Licensing Act 2003 which staggered closing times beyond the traditional 11pm cut-off.
The act – introduced in November 2005 – was based on the belief that staggered closing times would help avoid trouble as drinkers would not be leaving pubs and bars at the same time of night, so confrontations would be less likely, the study said.
But researchers said the act was built on “weak evidence that contradicted more credible and empirically-supported theories about alcohol availability and harm”.
The study used data from Greater Manchester Police to compare recorded rates of violence with licensed trading hours across the city from February 2004 to December 2007 – roughly two years either side of the change.
The researchers found no change in crime figures, with the increased staggering of closing times not associated with either lower or higher rates of violence.
They also found that opening times had increased by a far lower amount than anticipated - average trading times increased between 30 to 45 minutes per premise on weekdays and by one hour and 20 minutes at weekends.
Researcher David Humphreys said: "Over the past decade, England and Wales have witnessed a series of political prevention initiatives for alcohol-related harm that have been implemented largely without evaluation or systematic appraisal.
"This has resulted in missed opportunities to generate evidence and a missed opportunity to learn, both of and from, any mistakes."
He also suggested the recent implementation of a late-night levy in Newcastle would suffer from a similar lack of sound research.