The extent of crime in this country has been very much in the news over the last week or so.

The annual crime figures for England and Wales published at the Home Office on October 13 show a marked decline both in terms of offences recorded by the police and the results of the latest British Crime Survey.

Recorded crime has dropped 8 per cent over the year, while the BCS reports an overall decline of 14 per cent since the last survey which covered 1995.

The welcome news implied by the congruence of these two measures raises complex questions about the relative impact of criminal justice arrangements on crime alongside broader changes in society.These questions have been thrown into sharper relief with the publication by the United States Department of Justice of a special report, Crime and Justice in the United States and in England and Wales, 1 981-96.

This study is co-authored by Patrick Langan, a statistician with the Bureau of Justice Statistics and David Farrington, professor of clinical criminology at Cambridge University and currently a visiting fellow with the Justice Department.

Langan and Farrington have collaborated in this field for some years and parts of this study have featured in earlier publications.

The report was released in Washington on Sunday, October 4, but was not picked up by the British press until the last week-end.

As presented by some sections of the media here, the study confirms the simplistic "prison works" thesis and that America is reducing crime by increasing its use of imprisonment and that this is a road down which Britain should proceed.

This is not what the report stated and far less is it a valid conclusion to draw from its findings.In the first section of their report, Langan and Farrington use crime data as recorded by the police and the results of victim surveys to demonstrate that most types of crime declined in the United States over the period of the study and that by 1996 the rate (per inhabitants) of crime was lower than it was in England (all references to England include Wales).

The overall pattern in the United States has been one of decline in the early 1980s, rising thereafter until around 1993 and then once again declining.

In fact, the 1996 figures show most crimes to be at their lowest point over the period.

The trend in England, at least for half of the crime categories, was in the opposite direction with rates at their highest point over the period in 1995/96.The authors conclude from these data that Americans are in most respects safer from crime than people are in this country.

This conclusion, however, requires qualification in terms of crime category and type of measure.

For example, while both types of measure show the rate of robbery in England to have increased at a faster rate in England than in the United States, the respective difference across the two measures in 1995/96 are in opposite directions: the survey data show the English robbery rate as being 1.4 times higher than in the United States while as measured by police statistics the figure in the United States is 1.4 times that of England.

On the other hand, data for burglary show a consistent pattern of growth with the rate in England about double the American rate at the end of the period.

Despite some narrowing of the gap between the two countries, the rates for rape and homicide (the report refers to "murder" but an appendix confirms that the data relate to "homicide" which is a more comparable category across countries) remained 5.7 and 1.4 times higher, respectively, in the United States than for England over the period.

The homicide data are especially important because it is the fear of lethal violence which is so especially debilitating.

This point was forcefully made in a recent study (not cited by Langan and Farrington) by two scholars at the University of California, Franklin Zimring and Gordon Hawkins, who compared data on the homicides that resulted from victims of robbery and burglary in London and New York in 1992.In London some 212,000 incidents caused 7 deaths, while in New York 191,000 incidents caused 378 deaths, a total which is 54 times as great as in London (cited in Franklin Zimring, "Lethal Violence and the Overreach of American Imprisonment", National Institute of Justice, US Department of Justice, 1997).

It is fear of violent crime, especially when lethal, which is a potent aspect of the relatively tougher approach to cri me in the United States compared with this country and other parts of Europe.

One aspect of this, Zimring contends, is the overreach in the use of prison with respect to non-violent offenders.By attempting to compare data on offenders and their contact with the criminal justice process, Langan and Farrington begin by confirming the widely known finding that, with the exception of homicide, in both countries persons committing a crime are unlikely to be convicted.

For example, the likelihood of conviction for rape is about 20 per cent in the United States and about 10 per cent in England.

As the authors point out, these estimates are inflated as they are based not on the actual number of rapes but on the much lower number recorded by the police.

For burglary (where the base figure is based on victimisation data), the likelihood of conviction is around 1 per cent in both countries.

For robbery, assault and motor vehicle theft, the authors report that the likelihood of conviction in the United States was 2 per cent compared with 1 per cent in England.

They make the important point that these estimates refer to those offences being cleared-up and not to a particular offender being convicted.

Elsewhere, Farrington has found that 58 per cent of English males who admitted to committing burglary were eventually caught and convicted of burglary at least once before reaching age 33 (DP Farrington, "Self-reported and official offending from adolescence to adulthood" in MW Klein ed Cross-National Research in Self-reported Crime and Delinquency Kluwer, 1989).

But the point that only a small, and often very tiny, proportion of crime events result in a conviction in the courts underlines the marginality of criminal justice to the ups and downs in the level of crime.Despite its title, the study undertakes little exploration of how offenders move through the criminal justice process after the point of conviction.

Rather, the focus is the use of imprisonment by the courts.

The authors report that both countries imprison about 95 per cent of all persons convicted of murder.

For rape, the respective use of prison in England and the United States is 82 per cent and 95 per cent.

But for other offences, the United States is more inclined to resort to prison.

Taking burglary, for example, 60 per cent of convicted US burglars and 38 per cent of English convicted burglars were imprisoned; for assault, the figures were 62 per cent and 27 per cent respectively.

The authors' particular interest is the "risk of incarceration" which is reached by calculating the number of persons sentenced to imprisonment for a specific offence as a rate per 1000 of the population (the "incarceration rate") and then dividing that figure into the total number of victimisations for that offence.

This approach can be illustrated with reference to burglary.

The authors contend that the higher American incarceration rate for burglary cannot be attributed to a higher burglary incarceration rate as this is lower than the rate in England.

"Instead the higher US burglary victimisation rate is attributable to the higher rate at which the United States justice system catches, convicts and incarcerates burglars".The authors conclude that the data indicate that a burglary in the United States is more likely to lead to incarceration than one in England.

In other words, the risk of incarceration for burglary is four times greater in the US than in England.

Rather similar differences comparing the US to England were reported for robbery and assault.

Between 1981-1994, the risk of incarcerati on is said to have risen in the US whereas (with the exception of murder which remained stable), it declined in England.

In comparing time actually served in prison, the study shows that for every offence the US figures are substantially higher.

These additional time served data allow the construction of a "risk of punishment" measure, in other words the period of time spent in prison which an offender risks serving.

By this measure, the risk of punishment is generally greater in the United States than in England, and this risk is generally rising in the United States and falling in England.Searching for conclusions from this mass of data, Langan and Farrington claim that the patterns are somewhat clearer for England than for the United States.

They suggest that in England the declining conviction and incarceration rates do correlate with rising crime rates, but when time served data were taken into account the results become inconsistent.

With regards to the US, the trends in punishment and crime tend (the exception being burglary) lack consistency.

The implication drawn by the researchers is that "punishment trends and crime trends should not always be expected to have the same relationship in two countries over any period of time".

Furthermore, the authors are cautious to emphasise that even where correlations are established this does not establish causation and that crime rates may rise or fall for reasons that have nothing to do with the levels of punishment.

For example, with reference to the US, they calculate that demographic changes in the age and race composition might explain 41 per cent and 47 per cent of the drop in the US homicide and robbery rates, respectively, over the period and around one-fifth of the drop in the burglary rate.Can sensible policy implications be drawn from this all of this? A great deal, of course, depends upon the comparability of offence categories regarding the two measures which are used and across the two countries.

While there has been progress in terms of definitions across methods within both countries the same cannot be said of the situation across countries.

Difficulties also arise concerning efforts to compare differing processes of criminal justice.

As a recent Home Office report stated, with reference to an earlier version of the Langan and Farrington report, it is difficult to obtain comparability between measures and to disentangle sentencing from simultaneous events differentiating the areas studied (Reducing Offending, Home Office Research Study 187, 1998, p 94).

The Langan and Farrington study certainly does little to confirm the chorus of American "right realist" commentators such as John J DiIulio who contend that America is soft on crime or Charles Murray who periodically berate this country for not being tougher on crime.

Leaving to one side cross-national studies, even within a particular country, considerable caution is warranted.

Jeremy Travis, director of the National Institute of Justice which sponsored the Langan and Farrington study, while noting that the declines in violent crimes in the larger cities are greatest in those that have also experienced downward trends in the use of crack cocaine, observed how little is really known about declining crime rates in the United States.

His main point is that a real poverty of research design means it is not possible to conduct research in a way that reflects the complexity and potential power of real life and real communities.Almost as a post-script to their report, Langan and Farrington compare racial disparities in inc arceration across the two countries.

In the United States, the prison population rate (persons held on a given day in 1991) for adult whites was 396 per 100,000 compared with a rate of 2,563 for adult blacks.

In England, also for 1991, the prison population rates for whites and blacks were 102 and 667.

It follows that in both countries the prison population rate is approximately six times greater for blacks than it is for whites.

As Nils Christie, of the University of Oslo, has remarked in this context, there is no reason for European chauvinism vis a vis the United States (Nils Christie, Crime Control and Industry, Routledge, 1993, 122).