Don't distort statistics, Mr Espeut
Carol Coy, Guest Columnist
On Friday, November 7, 2014, The Gleaner published an article written by Mr Peter Espeut, titled 'How to lie with statistics'. Mr Espeut commented on the press release by the Statistical Institute of Jamaica (STATIN) on the findings from the July 2014 Labour Force Survey.
For more than 40 years STATIN has consistently presented data on the findings of the Labour Force Survey and adheres strictly to professional considerations in the dissemination of all its statistical products. It categorically denies any attempt to distort the findings of the Labour Force Survey by "seeking to paint a rosy picture" by not presenting all the data or by presenting "a superficial analysis".
The aim of the press release is to provide the main findings of the Labour Force Survey and not to give an analysis of the labour-market situation in Jamaica. Journalists and researchers who wish to undertake an analysis of the Labour Force Survey can access more detailed information on STATIN's website, www.statinja.gov.jm, or by contacting the institute.
STATIN would also like to correct major conceptual errors that were made in the article and hope that it will further enhance the reader's understanding of the data.
STATIN has been conducting the Labour Force Survey from 1968 using the concepts, definitions and methodology developed and established by the International Labour Organization, which allows for international comparability of the labour-market indicators that are produced by different countries.
Of note, an article explaining the concepts and definitions used in the labour force was published on behalf of STATIN by both The Gleaner and the Daily Observer newspapers to commemorate CARICOM Statistics Day on Wednesday, October 14, 2014. The main objective for publishing this article was to give the public a general understanding of the key concepts and definitions used to explain the labour market.
There are two important concepts that were reported inaccurately in Espeut's article. The first is related to the definition of the unemployed and the other to the reference week.
The unemployed are measured in two ways, using the 'strict' and the 'relaxed' definition. Using the strict definition, individuals are defined as unemployed if they are 14 years and over and, during the reference period, are 'without work', are 'currently available for work' and are 'seeking work'. Using this definition, all three conditions of 'without work', 'available for work' and 'seeking work' must be met simultaneously before a person can be considered to be unemployed.
For the relaxed definition, a person is defined as being unemployed if he or she is 'without work', and is 'currently available for work' even if no steps were taken to seek work.
The fundamental difference between the two definitions is that for the strict definition, one has to be seeking work to be classified as being unemployed, while for the relaxed definition, the individual just has to be available to work.
STATIN uses the relaxed definition of unemployment so all persons who are frustrated from seeking work, those who believe that there are no jobs available for them, and the general discouraged worker are classified as unemployed and are, therefore, included in the labour force.
Jamaica, like a number of other small developing countries, which have a large informal sector and do not have a developed system for registering the unemployed, uses the relaxed definition of unemployment.
If STATIN had used the strict definition, the unemployment rate for July 2014 would have been 8.4% instead of the official unemployment rate provided by STATIN of 13.8%.
The other conceptual inaccuracy in the article relates to the reference week for the survey. STATIN has a fixed reference week for each quarterly Labour Force Survey, which is the last full working week in the previous month before the start of data collection.
Consequently, all respondents are asked questions about their economic activities in respect of that particular week. The article implies that the survey reference week is the week previous to being visited by the interviewers and would therefore vary according to the date of the interviewer's visit.
STATIN accepts the fact that the measure of the labour market is very complex and welcomes discussions and analyses from researchers and journalists that can further enhance readers' understanding of the data. We, however, expect that an accurate representation of the methodology used by STATIN will be provided in these discussions.
Carol Coy is director general of STATIN. Email feedback to email@example.com.