The new statistical techniques used to identify the problem could be employed in the financial markets to identify patterns of insider trading activity, the researchers say in the January issue of the Economic Journal.
Using an analysis of betting data on 5,000 horses during the 1992 flat racing season, Leighton Vaughan Williams and David Paton of Nottingham Trent University have identified insider trading patterns by looking at the behaviour of bookmakers in setting odds.
To avoid being stung, bookies shorten the odds on horses - and raise their profit margins - on certain types of race where they suspect insider knowledge is being used by punters.
This applies particularly to long-shot horses in little known races with large numbers of entrants, where bookmakers are most fearful of insider activity.
Only in high-class handicap races, which are subject to intense media scrutiny, are the researchers unable to identify any insider trading effects at all.
The ESRC said the research came soon after the Jockey Club admitted the industry was rife with corruption and race-fixing and that some jockeys were receiving sexual favours in return for inside information. All punters feel the effects of insider trading because of the bookies' reaction in reducing odds, especially on long-shots, the paper said.
The researchers added: "Not all insider trading is illegal in betting circles. Even so, these findings can only provide further ammunition to those seeking to crack down on illegal practices in the racing industry."