Driven by data; ridden with liberty.
The desire for partisan explanations often trumps more considerate forms of data analysis. This is particularly true in the way that offical statistics data sources, such as the Office for National Statistics and the Health and Social Care Information Centre (HSCIC), are interpreted.
Dr Gabriel Scally, a “public health doctor” and trained general practitioner, asked about the declining rates of cervical screening coverage in England:
Women’s health at risk from dropping cervical screen rates in England. Is [public health] failing due to Tory cuts?
The chart poses three immediate questions: why does the chart begin at 72.5%, enlarging the perception of the fall? Why is a time series being represented by a histogram? Why is no data prior to 2011 provided?
Dr Scally did manage to answer the third question: a new measure of coverage was calculated, where age appropriateness is taken into account, meaning it was between 3.5 to 5.5 years since the last adequate test. This new measure began in 2011. The former measure is simply that it is less than 5 years since the last adequate test.
Dr Scally also replied that “government statistics routinely use a truncated axis”. This varies by report, and is true for the HSCIC report  that the cervical screening coverage data is drawn from:
This chart’s axis is also truncated, but begins at 65%. Moreover, this report includes the former measure of cervical screening coverage for comparison.
This measure has been falling slowly and irregularly since 2005. This series rather belies the notion it is “Tory cuts” causing the decline, rather than a lack of propagation under successive governments.
 HSCIC, 2015. Cervical Screening Programme, England. Statistics for 2014-15. Available from: http://www.hscic.gov.uk/catalogue/PUB18932/nhs-cervical-stat-eng-2014-15-rep.pdf [Accessed: 23rd November 2015]