TT Stats Director explains CSO data issues

The following statement, issued by the Ministry of Planning and Development and authored by Sean O’Brien, Director of Statistics (Ag), was sent to the media today.

Principle four (4) of the United Nations Fundamental Principles of Official Statistics (UNFPOS) states “The statistical agencies are entitled to comment on erroneous interpretation and misuse of statistics”. In this context, I feel compelled to respond to the plethora of recent comments and criticisms from diverse national commentators regarding the data produced by the CSO, which in my opinion emanate from erroneous interpretations of these data.

The National Unemployment Rate in particular has been the target of severe condemnation from many of our stakeholders such as the trade unions, academics, journalists, economists and politicians to name a few. The central thesis of these complaints is that the unemployment rate simply does not reflect the numerous, media reports of retrenchments, business closures and other job losses experienced by the economy.

From my perspective, the problem here is that an intuitive understanding of the unemployment rate is an insufficient basis on which competent conclusions can be proffered regarding this complex statistical indicator, which is influenced by several factors.

To lend clarity to the issue it must be understood that the CSO is an adherent to the International Labour Organisation’s Standards and Guidelines on Labour Statistics. Accordingly, the definition of an unemployed person is someone who has not worked during the period of enumeration, but is willing and able to work and has actively sought work during a specified reference period.

Therefore, persons who have suffered retrenchments and other job losses, and have not sought re-employment during the reference period; must not be classified as unemployed, contrary to the supposition of many commentators.

Further, persons without jobs who are not seeking employment are not to be considered as part of the labour force in accordance with the ILO’s Standards and Guidelines. It should be noted that the unemployment rate is calculated by dividing the total unemployed by the total labour force and multiplying the result by one hundred.

Therefore, to the extent that persons who have lost jobs do not report that they have looked for employment; then the denominator (the labour force) falls and the numerator (the unemployed) remains unaffected thereby putting downward pressure on the unemployment rate.

By way of example, the CSO’s 2016 Fourth Quarter Labour Bulletin reveals that there were five thousand, two hundred (5,200) job losses when compared with the third quarter of 2016. However, the unemployment rate fell from 4% in the third quarter to 3.6% in the fourth quarter. Notably the data indicate that the labour force fell by seven thousand, nine hundred (7,900) persons.

Therefore, the significant decrease in the number of persons with jobs was accompanied by a decrease in the unemployment rate mainly due to a sharp fall in the labour force. This serves to illustrate that it is misleading to comment on the unemployment rate in isolation of the other available labour data.

Critics of CSO’s labour data have also tended to make comparisons of the unemployment rate of the quarter currently under review with that of the quarter that immediately precedes it. This practice ignores the element of seasonality that exists in the labour market.

To better understand the labour force’s characteristics it is useful to compare the fourth quarter of 2016 with that of 2015 so as to ensure that changes in the unemployment rate are not simply due to seasonality. Further, the CSO does not record unemployment data from firms and other establishments as seemingly assumed by many. Labour data are gleaned by enumeration of the households via the conduct of the Continuous Sample Survey of Population (CSSP).

In this manner, persons who have been retrenched from establishments but have found other jobs either in the formal or informal sector can be recorded accordingly. Given that the firms do not report on the current status of their former employees it is sophomoric to connect reports of job losses at establishments with an equal rise in the unemployment rate.

Beyond the labour statistics, many in the media and general public continue to lament that there are many datasets that have not been updated by the CSO. The fact of the matter is that much of the data that are not available from the CSO are simply not available to the organisation. It appears to be a little known fact that the CSO depends on the various government ministries and departments to provide it with the source data which are then processed and disseminated by the organisation.

In fact, the ensemble of these ministries and agencies in concert with the National Statistical Office is known in the literature as the National Statistical System (NSS). In Trinidad and Tobago the NSS is not functioning well, given that many ministries are struggling to prepare the source data at the requisite level of quality.

In fairness to the respective ministries, statistical reporting is not their core function. The CSO is actually mandated to lend statistical mentoring to these bodies and to generally coordinate the NSS, but at present, as a mere department of the Ministry of Planning and Development it lacks the authority to do so. In fact the CSO is prohibited from collecting key data, (most notably from the V.A.T. Office).

In these circumstances the CSO is to be transitioned to the National Statistical Institute of Trinidad and Tobago (NSITT) in the near future which is envisioned to be enabled by strong, contemporary statistical legislation. The Management of the CSO and the executive of the Ministry of Planning and Development are working so that the new NSITT will be empowered to coordinate the NSS and to collect all the necessary data and will have the legislative potency that its predecessor lacked.

In conclusion, the CSO does not wish to discourage constructive criticism of its statistical products and the attendant methodologies. In fact the organisation has always been comprehensively peer reviewed by several regional and international development partners and has made many methodological improvements based on their findings.

However criticisms based on a cursory understanding of the data and the NSS, serve only to unnecessarily erode confidence in the CSO and its data which leads the discouragement of national and foreign investment and a higher cost of borrowing internationally.

The Central Statistical Office falls under the purview of the Ministry of Planning and Development.