Pulling a story out of third-party data has a slew of advantages. One of the most useful is that a large chunk of the information gathering has been done for you by an organisation that often has resources or access that is beyond most journalists. The story is already there – you just have to sift through facts and figures to find it.
But this helping hand has a huge downside. Because someone else has found your information for you, they have decided what they want to find out, and as often as not that means they have not zeroed in on the data that makes the best story.
The statistics on average weekly earnings from the Office for National Statistics (ONS) provide a frustrating example. This data is fascinating because it allows you to look at how much workers in different areas of the economy are earning each week, and how those figures change each month.
Yet the categories it is divided up into do not make it easy to pick out the best story thread. Given the current debate around bankerss bonuses, the most interesting figures concerned earnings in the financial services sector. Yet of the two different catagorisations that included banks and other high-paid financial services workers such as hedge-fund managers or private equity managers, neither allowed you to completely separate out other jobs that to the public, and journalists, seem very different.
The first set of categories (by sector), separates out “financial and business services”. But this includes not only all workers in banks, but a whole range of other jobs such as real estate agents and even grounds keepers. The trends in earnings are interesting, but each statement on who is actually seeing their earnings go up, and what this means, has to be hugely qualified.
The second set of categories (by industry) was more useful as it narrowed down the data that includes the finance sector that has dominated the news over the last two years into “Finance and insurance activities”. Yet though this provides a more specific measure, and offered a more striking set of data, it still needs constant qualification to point out that, while banking and similar activities form a huge chunk of this measure, they are not the whole story.
Things got even harder when trying to cross reference this data with Gross Domestic Product (GDP). The ONS doesn’t produce GDP growth figures by industry. This means that while some comparison between average earnings in a sector and overall growth in its output is possible, the news point was hugely diluted by the inability to pin down the workers, and salaries, that were most interesting.
The data on average earnings paints a fascinating picture of how workers in different parts of the economy are rewarded. It is data that would be almost impossible for any journalist, or team of journalists, to collect.
But because the data was not collected for the purpose of generating an attention-grabbing headline, a journalist must pick out, explain, qualify and contextualize the data throughly to maintain accuracy and help a reader make sense of the numbers. Furthermore, its use is in many ways more limited than the sort of information gathered by a journalist asking exactly the questions he wants the answers to.
This is a trade-off that happens time and again when trying to take data gathered elsewhere and turn it into something that makes good news.