Virtually fact-free

When the state of Georgia executed Troy Davis last week, everyone was outraged – one way or another. The liberal left were outraged that a man whose conviction rested on nine witness testimonies was denied clemency when seven of them were withdrawn. The conservative right were outraged that anyone would question the exalted justice system of God’s own country, or the wisdom of forever preventing a convicted murderer doing harm. The libertarian right were outraged, simply, at a government which has the audacity to take away human life.

I was outraged about something else. At the end of last week, Priti Patel, a UK Member of Parliament for the Conservative Party, went on the popular British panel show Question Time to debate, amongst other things, the case for death penalty, in light of Davis’ execution. In her spirited defence of capital punishment, Ms Patel said many things I, and many others, disagreed with. This post focuses on one in particular, and the huge problem it underlined.

A fellow panel member asked Ms Patel whether there was any evidence that the death penalty actually deters people from committing murders. “There is evidence,” was her reply. And that was that.

Now, a panel show is a difficult medium in which to qualify statements, and Ms Patel’s willingness to commit herself is a refreshing quality in a British politician. But the question remains: what evidence?

I’ll be damned if I can find it. I am a man of my generation, so I Googled for answers. A search for ‘death penalty deterrent’ brings up mostly campaign groups, perhaps unsurprisingly. Amnesty, the human rights campaign group, presents charts showing murder rates in US states which do and do not use the death penalty. Rates tend to be far lower in states which abstain, but  conflating factors - poverty, for example – render this data fairly meaningless.

Prodeathpenalty.com  – a website wearing its heart on its sleeve, if ever there was one – hosts an article by George E. Pataki, the former Governer of New York who ran on a fierce pro-death penalty ticket. His essay is worth a read, though again the simple picture he paints is far from the truth - Freakonomics, Steven D. Levitt’s bestselling economics book, illustrates the problems with inferring anything very much from US crime figures.

Further searching reveals one credible, modern source, again based on US data (the only large, Western country which has carried out significant numbers of executions in modern times). The source is a survey of leading criminologists from 2009, published in The Journal of Criminal Law & Criminology. Their conclusion is clear:

“The findings demonstrate an overwhelming consensus among these criminologists that the empirical research conducted on the deterrence question strongly supports the conclusion that the death penalty does not add deterrent effects to those already achieved by long imprisonment.”

My point isn’t that Ms Patel’s assertion was wrong, but that her source was not immediately obvious. She may well be basing her claim on very good data; she is, after all,  an intelligent woman. But we have no idea what that data is. Her statement was indicative of a huge problem in journalism and politics – the tendency to quote ‘facts’ without context or source. It is not a new problem, but there is a new solution: the Internet.

But firstly, a slightly different example.

 

The 6.5% rape claim

It is undoubtedly an arresting figure: between 2007 and 2008, 6.5% of reported rapes result in a rape conviction. It has been bandied around to support all kinds of things, by all sorts of people. It is the crime and punishment equivalent of Katie Price, generating column inches at an improbable rate (and often misunderstood?). So what does it mean?

As it happens, The Guardian recently wrote an in-depth piece on this very statistic. The number itself is real enough: 6.5% of reported rapes result in a rape conviction. Only 6.5% – and I feel that that number does deserve an ‘only’. But this simple fact is chronically misunderstood.

13% of reported rapes results in a conviction of some form – half (6.5%) result in a rape conviction, and the rest are for lesser charges, such as sexual assault. The 6.5% figure has been contrasted with the conviction rate for other violent crimes - such as violence against the person, which between 2002 and 2006 stood at 70% in England and Wales. But this comparison is false: the rape number is from initial report to final conviction, whereas the ‘conviction rate’, as defined by the Home Office, concerns only cases going to trial. The comparable figure for rape, in 2007, was 47%.

There is still a gulf here. It is absolutely fair to say that rape conviction rates are too low. Further, the misinformation which rape apologists seed is faintly sickening: notwithstanding some high profile exceptions, false accusations of rape are actually extremely low.

Some have suggested that this is a peculiarly British problem – since 6.5% is far lower than the comparable 25% figure for France. But, again, this is misleading – rapes are much more often reported in Britain. Countries with higher reporting rates tend to have worse conviction rates. If anything, our high reporting rate should be congratulated.

Why does this matter? I would argue (though I have no evidence) that this single figure is extremely damaging: a woman who thinks rapes are ‘unconvictable’ is a woman less likely to report a rape. Ironically, if this true, we can expect an increase in the 6.5% figure, as less rapes are reported. Is this really a victory? Should we not, instead, question the high ‘attrition’ rate in reported cases, and ask why the Police often fail to provide rape kits to victims?

So rape statistics are another case where proper referencing and proper clarity would be incredibly valuable. If everyone understood where these figures came from, we could be achieving more to prevent rapes.

 

Whose line is it anyway?

As Homer Simpson put it: “People can come up with statistics to prove anything.” Mostly, however, people don’t do so wilfully. Rather, facts are rephrased, rehashed, echoed. The original sources are obscured, context is removed, and much more weight is put on them than they can bear.

Journalists are of course hampered in efforts to properly reference their stories by basic practicalities: newspapers don’t have room, and broadcast media don’t have time. Politicians suffer similar constraints: a snappy political soundbite loses its snappiness when referenced. Acknowledging sources may also detract from a good yarn; I’m sure Johann Hari would agree.

But there is a third way: if journalists and if politicians are sincere in their pledges of transparency (and there is now an unprecedented public appetite for them to be so), they can provide sources for their claims online. Every article in The Times or the Daily Mail can and should be referenced online. Every ministerial speech should appear in referenced form on DirectGov. The Internet provides unlimited space and, in modern Britain, near-universal access.

This is not a criticism of opinion journalism; long may our proud tradition of mouthing off continue. From Jan Moir to Polly Toynbee, opinion journalists enrich public debate in this country. Nor does it preclude anonymous sources. Rather, it is an appeal: make it clear where any stated facts come from.

Some journalists – notably The Guardian’s George Monbiot and blogger Ben Goldacre – are already beating this drum. The Guardian and The Telegraph have partially taken up the mantle – their online comment pieces are hyperlinked to sources, albeit incompletely. A reference pack for their daily paper editions, however, has not been forthcoming.

Let’s put it another way: it is curious to me that we still consider The Times as a ‘more valid’ source than Wikipedia. Why? Both have their inaccuracies, but at least Wikipedia is referenced.

 

Standard form

A slightly-mysterious group of people in Switzerland, known collectively as the International Organisation for Standardizaiton, could help. They enforce internationally-agreed standards on all sorts of things, from environmental waste management to photographic film. They accredit organisations which comply with their standards. If you work in a large company, it will almost certainly have accreditations in quality control and environmental responsibility.

As of yet, there is no agreed ISO standard for the media to adhere to: a standard that demands all articles are referenced online, and that employed journalists have declared their interests; a standard that requires all staff to be trained in the legality of phone hacking, and to post full transcripts of interviews. There should be.

Political parties should also seek ISO accreditation. MPs’ pay should be docked when they stray. Press passes should be denied to any media organisation that falls short of agreed standards.

Times have changed, and what journalists and politicians say is no longer taken on faith. At the moment, it isn’t taken on anything; trust in our politicians and our journalists has never been lower (and note that the linked research precedes the News of the World hacking scandal).

Well, if you can’t be trustworthy, you can at least be transparent. We can and should demand more from opinion leaders in our country. Where do your opinions come from? Are they based on fact, hunch, or prejudice? If they can’t tell us, then their opinions are worth less than the paper they are written on.