Charities can bend the truth, too

Statistics quoted by charities should always be checked. As organisations they may seem too well-meaning to misrepresent anything, but a charity’s income depends on making a problem appear serious and growing.
For example, last week the charity Refuge chose to “reveal” that the number of women killed in domestic violence incidents had increased from 72 in 2008 to 101 in 2009. Was this true, and is so, was it meaningful?
The actual data since 1998-99, tabulated below, show the numbers of women killed by a partner or ex-partner for every year since 1998-99:

This makes it clear that although there was indeed an increase between 2007-08 and 2008-09 in the number of women killed by partners or ex-partners, there is no clear trend. Totals fluctuate from year to year, have several times been higher than 101 in recent years, and were not 72 in 2007-08, but 78.
The statistics also show that fewer women in total were killed in 2008-09 than in any year back as far as 1998-99. (Ignore 2002-03, distorted by murders attributed to Harold Shipman, and bear in mind that both of these 2008-09 figures may be revised upwards as the data available when the statistics were compiled ran only to November 24 2009.)
Refuge’s figures were quoted without any reservations by The Guardian, The Standard, Metro, and the Daily Mirror, and probably left most readers of those papers with the impression that this problem is getting worse year by year. In fact, it isn't – and the figures for domestic violence generally have declined significantly since 2004-05.
Another charity dipping its toe in muddy waters was Crisis, which seized on last week's homelessness statistics, expressing alarm at a rise in the third quarter of 2010. This could be the end of a long-term downward trend in homelessness figures, though it’s too soon to say. Table 2 shows the numbers of applicants accepted as homeless in the third quarter of the past eight years, and do show a small increase between 2009 and 2010, though to nothing like the levels seen just a few years ago.

So was Crisis right to express alarm and say it could be “just the start of something bigger”? Maybe, but the report in The Guardian made no attempt to put the claim in context, and claimed that the figures represented those “sleeping on the streets”.
It does not, of course, as the paper made clear in a correction today; many of the claimants are living in overcrowded conditions, sharing with friends, or have recently suffered the breakdown of a relationship. They aren’t necessarily on the streets.
It’s Christmas, when all the charities hope to be remembered. So the temptation to exaggerate and get their names in the papers is greater than usual. It seems to work.