No News is Good News
So, according to the Cambridge News, in one year, Cambridgeshire Police managed to reduce its monthly 'translation' bill for Romanian and Bulgarian from £1,366.94 in January 2013 to £0 in January 2014. However, the word 'translation' here must refer to 'written translation' not '[verbal] translation' or 'interpreting'. This is proved by the fact that, although January 2014 might have been a very quiet month for interpreters in Cambridgeshire, there were nevertheless interpreting assignments for which - obviously - interpreters were paid more than £0. Furthermore, amongst police forces, written translations make up only a small part of the overall expenditure on interpreting and translation. This is obvious from the minute expenditure on Bulgarian 'written translations' in January 2013, namely £9.10, a sum, incidentally, so meagre as to be below any reasonable minimum charge. In other words 'not very much expenditure' in January 2013, has been reduced to 'nil' in January 2014.
This is good news for Cambridge MP Julian Huppert who has a gift for stating the obvious. He notes that 'these statistics [show] that Cambridgeshire Police have not had to employ the services of Bulgarian & Romanian translators very often'. However the MP soon slides into a generalisation about the costs of 'translation services'. He observes that that 'over the last year' there were few demands on the Police budget 'for translation services'. He seems to think that the statistics arising from a comparison between one January and another involving only two languages can be extrapolated across all languages over an entire year. Huppert seems anxious to assure the folk of Cambridgeshire that far from there being a sudden influx of Balkan peoples in January 2014, the whole of the preceding year had itself been a very quiet one for immigration in general. In Huppert's world, all has been well for Cambridgeshire tax payers for some time. But it is entirely possible that other languages such as Polish & Lithuanian were all-too-much in demand in 2013, especially as those nationalities could enter the UK free from the restrictions imposed on Romanians and Bulgarians.
Extrapolating statistics is one thing, but using figures relating to 'written [police] translations' as a barometer of the overall rate of immigration is absurd. As we have seen, no police force ever has to use written translators 'very often'. If the Cambridge Evening News and Mr Huppert were really interested in expenditure on 'written translation' they would have asked local councils or the NHS or even translation agencies how much is spent on such documents per month. In any case comparing the two months of January 2013 & 2014 will tell you nothing. The amount spent on Romanian written translations in January 2013 (£1,357.84) may have been spent on a handful of one-off official documents which will have skewed that January's figures. The fact that the corresponding figure for this January was £0 would tend to support this argument. In other words whole months may pass in Cambridgeshire without the police having any requirement for written translation in certain languages. Even if one were to compare the interpreting or 'spoken translation' costs incurred in January 2013 and January 2014, one would find the comparison unhelpful. Cambridgeshire Police deal in crime. Crime is affected by the weather. This January the rain was unrelenting. In such conditions most criminals stay at home. Foreign criminals meanwhile are no different to their law-abiding countrymen in their tendency to enjoy an extended Christmas in their country of origin. In other words, in January 2014 such criminals had either stayed at home in the UK and kept dry, or enjoyed a protracted New Year's holiday with their relatives in Bulgaria and Romania. In any case, the Romanians and Bulgarians who were expected to arrive in numbers in January 2014 will have been law-abiding job-seekers. Self-evidently the relaxation of work restrictions will have led to an influx of workers. Their arrival will not have greatly affected the statistics relating to crime.
The article should have made it clear that 'spoken translation' or 'interpreting' makes up the bulk of Cambridgeshire Police's requests for language services. The force's figures for interpreting expenditure for the calendar years 2009 and 2010 are available. These show that the average annual expenditure over those two years on interpreting was £645,492.30. The Cambridge News article suggests that the 2009/2010 annual figure for both interpreting and translation was £809,000. Taken together these statistics suggest that interpreting, or 'verbal translation' accounts for almost 80% of the combined interpreting and translation bill.
Yet we also learn from the article that Bulgarian and Romanian interpreters cost Cambridgeshire Police 'a tiny percentage of the total bill which stretches into three figures'. Here the article has switched its focus from a comparison between two successive Januaries to a consideration of the annual figures. More importantly, it has also switched its focus to interpreting costs. For a moment it seems that the article is about to address the crux of the statistical issue. However what the 'total bill' refers to is not clearly defined. One assumes that it reflects the aggregate annual cost of interpreting and translating in all languages. If so, Romanian and Bulgarian interpreting (minus translating) will account for a misleadingly diminished percentage of that total bill. The article should have calculated the combined Romanian and Bulgarian interpreting and translating costs as a percentage of this 'total bill'. That however would have made the 'tiny percentage' seem rather more than 'tiny'.
The article tells us that in 2006/2007 in Cambridgeshire interpreting or translating services were required for 100 languages. Meanwhile the 'total bill' for language services in 2009/2010 was, as we have seen, £809,000. If we assume all 100 languages continue to be in demand, we find that an average language accounted for £8,090 - or 1% - of that year's bill. In fact between 2009 and 2011, as a percentage of the total [interpreting] bill, Romanian and Bulgarian interpreting costs increased from 3.09% to 4.28% to 5.17%. Furthermore, between 2009 and 2011 Romanian came 10th 8th and then 7th in the league table of expenditure on interpreting. Over the same period meanwhile, Bulgarian came 27th 28th and 20th. Clearly, over this period, interpreters in these languages were ever more in demand, and the costs related to these languages rose incrementally. And, as we have seen, far from being average languages in 2011 accounting for a nominal 2% of the interpreting budget, Bulgarian & Romanian accounted for 5.17%. One set of irrelevant figures relating to written translation work in two successive Januaries will not disguise the annual increase in demand for Bulgarian and Romanian interpreters.
In general, the words 'tiny percentage' are misleading. For how exactly does one define a 'tiny percentage'. Does it refer to a single digit percentage? Or just 2-3%? In fact as we have seen the most recent year for which figures are available prove that the figure for Romanian and Bulgarian interpreting accounts for 5.17% of interpreting costs. As we have also seen, this percentage figure will have to fall if translating costs are (misleadingly) included in the total bill. But even 4% would represent a sizeable chunk of the force's overall six-figure costs. In any case, if one juxtaposes the phrase 'tiny percentage' with figures such £1,357.84 or £9.10 one cannot be surprised if the casual reader forms the impression that the 'tiny percentage' itself relates to something akin to 'loose change'.
The article has two clear goals. It is concerned to allay fears over (a) a sudden influx of Romanians and Bulgarians into the UK and (b) the resulting increased pressure on the public purse. It does all it can to squeeze the results it wants from whatever figures are at its disposal. The quality of its statistical analysis is embarrassing. What is most embarrassing however is the article's determination to fudge the terms 'translation' and 'interpreting'. Examples abound. In one paragraph we learn that 'the bill for [Cambridgeshire Police] force's translators has fallen by a half'. Elsewhere we discover that the 'translation bill' for 2006/7 was 'at least £800,000' and that 'the cost of translators was £677,000 in 2008/9'. All these figures relate to the combined interpreting and translation costs incurred by the force. Yet many readers will assume the words 'translators' and 'translation' refer to the 'written translators' and 'written translations' referred to earlier. They will therefore form the erroneous impression that six-figure costs have been [very recently] slashed to almost nothing, given that Bulgarian and Romanian 'translation bill' for January 2014 is 'zero'. Indeed even when Huppert refers to the low demand for 'translation services' in 2013 we cannot be sure that the casual reader will not take this phrase to include 'interpreting'.
Similarly the article suggests that the top seven 'most commonly translated languages' include Romanian and Bulgarian. Yet, as we have seen from the statistics available up to 2011, the Bulgarian language was far from being a major player in 'interpreting' terms. It was nowhere near being a top seven language. Romanian meanwhile just scraped into the top seven in 2011. Or does the phrase 'the most commonly translated languages' refer instead to a much more recent year? If so it seems there may be much to worry about from an influx of Romanians and Bulgarians. However, we should remind ourselves that we have been told that Romanian and Bulgarian interpreting costs constitute a 'tiny percentage' of the overall bill. Presumably then the phrase 'the most commonly translated languages' here must instead refer to the 'the [languages] most commonly translated in writing'. Unfortunately the casual reader will then reason that the other 93 - less popular - languages must be costing the force almost nothing, given how cheap the Romanian and Bulgarian 'translation bill' has recently become.
In summary, the article resolutely refuses to define its terms clearly. It uses 'translation' to mean 'written translation' at some points, whilst in other places the word clearly means 'translation and interpreting'. Yet 'interpreting' is also distinguished from 'translation' when it suits the article's purpose. This approach impoverishes the intellectual rigour of the article which uses the flimsiest of evidence and the confusion between the words 'interpreting' and 'translating' to give the reader the latitude to draw conclusions that are unjustified. Furthermore, the article cannot avoid being contradictory even when the (confused) reader restricts word 'translation' to its specific meaning of 'written translation'. For instance the article maintains that Romanian and Bulgarian are among the top seven 'most commonly translated languages' whilst Huppert maintains that the police 'have not had to employ the services of Bulgarian and Romanian translators very often'. These two positions are mutually contradictory.
However it is touching it is that an FOI request has been graced with such a rapid response. Interpreters can only dream of receiveing such prompt attention to their FOI requests. It is also however unsurprising that this FOI response has been 'spun' to paint a rosy picture of falling immigration rates and the consequent savings to the state. The truth about FOIs is that they have become a specious tool by which politicians seek to justify and bolster their preconceived agenda whilst all the time appearing oh-so-democratic and truth-seeking. The actual truth is that there are no general conclusions that can be drawn from a consideration of statistics relating to two Januaries' worth of police costs incurred on written translations. Anyone who thinks otherwise is a politician.
Lastly, the article reports the words of a police spokesman who observes that language requests are managed 'in a way that minimises waiting times and saves money'. How odd that exactly the same words were expressed by a Cambridgeshire Police spokesman on April 13th 2013 in an article written by the same journalist in the same paper. This earlier article also contains a large number of other coincidences with the current article. Either this journalist has cut and pasted parts of his earlier article to flesh out what is a wafer-thin 'story' or the police spokesman has trotted out the same comments in response to similar questions but in a different context. Revealingly, it was the earlier article which disclosed the fact that Romanian and Bulgarian were among the top seven requested languages. Now we know (finally) why this current article is so contradictory. It is a pastiche which has drawn inspiration from 'facts' that are nearly a year old. These regurgitated 'facts' have been parachuted into an article that is based on extremely specific and very recent statistics with which those 'facts' have no effective relationship. No wonder there is confusion. But what the current article does not repeat is the large sums spent by the county council on 'translators' since 2009, namely £300,000. However to repeat that statistic in the current article would have severely undermined the point being made by the apparent collapse in 'translator' costs in January 2014. In other words, whilst the costs of processing foreign criminals may be falling, that does not mean that state financial support for immigrants is not running - and will not continue to run - at an all-time high.
Now the words 'waiting times' used by this spokesman clearly reveal that he or she is thinking only of interpreting. As distinct from human interpreters, written translations do not 'wait' in police stations. This spokesman clearly does not imagine 'written translation' constitutes an important element in cost-savings otherwise he or she would have mentioned it. Ironically this provides the clearest proof possible that 'written translation' costs are of minimal consequence to the police budget. This undermines the entire thrust of the article, and reveals yet again the dangers inherent in pastiche journalism.
In general, this police spokesman would like the reader to think that the fall in translation costs is largely due to efficient management of interpreter time. It is true that there have been major improvements in this area, and interpreters now rarely spend more than two hours at a Cambridgeshire police station for a single detainee. However the enormous savings in 'translation costs' made between 2009/10 and 20010/11 (from £809,000 down to £420,000) are due in the main to a circumstance that the spokesman does not seem anxious to discuss. In October 2010 swingeing cuts were made to interpreter pay by Cambridgeshire Police. A not untypical short, early evening job for which an interpreter had previously earned £138.25 was reduced to £60, a fall of 56.6%. Many longer weekend jobs fell by 40%. It is no coincidence therefore that the 48% savings made by Cambridgeshire Police between 2009/10 and 2010/11 is also the average of 56.6% and 40%. Cambridgeshire Police seem to have realised that the period between 8pm and 10pm was a particularly busy period for foreign language interviews. They therefore changed the start of the evening (or 'unsocial hours') period from 8pm to 10pm. They also decided that if the interpreter was not on the premises for the full two-hour minimum charge period (for which £60 was guaranteed), any time that remained unused should be deducted from the travel time. Thus if a job lasted 1.5 hours (£45) and the travel time lasted an hour (£15), the interpreter would get the minimum charge (£60) with no expenses for travel. In other words Cambridgeshire Police carefully calibrated the terms and conditions of interpreter assignments in order to complement their reductions in hourly pay. Yet nowhere is there any mention of this in either of this reporter's two articles. This is apparently what passes for journalism at the Cambridge News.