CSO key to evidence-based policymaking

DR ROGER HOSEIN

The idea of “evidence-based policymaking” has gained ascendency in the global economy and indeed the management of the covid19 situation was heavily reliant, where it could, on data. Policy decisions in the increasingly complex global economy should be informed by careful analysis using data that is sound and transparent. Evidence-based policymaking must be the way forward for all economies in making decisions to recognise issues of national importance and in so doing to formulate important solutions.

Indeed when decisions are made on criteria other than sound evidence then it can promote distorted outcomes. Such biased decision making may be done to satisfy pressure groups of relevance to political gain or for rent seeking purposes with kickbacks.

Among researchers it is generally accepted that evidence-based policymaking helps support resource allocation decisions by the State and other groups to give the best outcomes from a democratic perspective as it is characterised by the principles of fairness and transparency. Formal econometric and statistical assessments of the economy depend heavily on models that require a wide block of time series data. Models on the labour market, for example, use detailed time series data sets to understand the behaviour of the unemployment rate, the labour force participation rate, the wage rate, and other related variables.

In this regard, a vibrant Central Statistical Office (CSO) producing timely and reliable data is central to the planning process in this economy. One of the flagship publications by the CSO in my judgement is the Labour Force Report, which uses data from the Continuous Sample Survey of the Population (CSSP). The CSSP provides a block of valuable data on the economy. It is designed as a multi-purpose household survey (it started in 1963) and has as its objective to provide up-to-date data on labour force characteristics on a continuous basis.

The CSSP’s survey design has had several revisions, with the latest in 1994. The data in the CSSP is based on a stratified cluster design and covers data on the labour force by:

(a) age

(b) administrative area

(c) gender

(d) employment status

(e) educational attainment

(f) occupational group

(g) industrial group

(h) industrial sector

(i) type of worker

Data from the CSSP is also available on people with jobs and unemployed people, new entrants into the labour force and income group. The last CSSP block of data was in 2016. This makes it difficult to line up an understanding of the structural changes that took place in the economy during the depression and it is something that we need to fix quickly moving forward.

In this regard I ask: how do you build your economic recovery plan using such outdated data? Do we assume that the labour market in 2020 is the same as in 2016, say, by occupational group or by industrial sector? Policymakers must not get accustomed to making decisions using outdated data as you may make decisions that do not fit the reality of the moment. Let me use an example: if we were to use the most up-to-date unemployment data from the CSO – mid-2018 – the unemployment rate is 4.8 per cent.

However, the Petrotrin refinery was closed in late 2018 and the static and ripple effects of that data may have pushed the unemployment rate to say eight per cent. Going into covid19 planning that type of data would be critical as the unemployed in many cases will be closer to the poverty line than other groups within the labour force.

Running the country on knee-jerk actions with limited data is akin to hit and miss in T20 cricket, sometimes with your eyes closed. Policy outcomes in such a setting can be mixed. Let us look at the Gasparillo area. Even if we were to use data at the level of the county, it is woefully outdated so we cannot as yet tell the impact of the closure of the refinery on the community using hard core data. How, then, do you plan? How do you know if the educational attainment in the Gasparillo area could sustain an eTecK park within the community?

Three immediate pieces of data that may add good value

Three specific pieces of data that I think will be urgently needed going forward. Firstly, the CSSP provides data on the labour force and the review of the economy provides data on output. To calculate output per worker we need an alignment between the way the data is provided by the CSSP in conjunction with the review of the economy’s revision for GDP data. The CSO will need to get its act together quickly in providing this data so that policy makers and researchers will have an idea of where the highest output per worker exists, because output per worker can in turn influence the incentive structure and hence the eventual output level in an economy.

The labour force participation rate is a critical variable in terms of understanding the gap between the labour force and the non-institutionalised population. Without up-to-date data perhaps at most one quarter or two quarters old, it will be difficult for policymakers to make sound judgements as to how to improve the labour force participation rate and what are the likely determining and influencing factors so that the downward trend in this variable could be reversed.

Finally, labour force data by type of worker is critical as the TT economy is in a depression (not a recession) since 2016 and by all indicators this will be accentuated in 2020. As a consequence labour force data by type of worker would help policy makers to understand what the size of the private sector is as well as the distribution of employment in the private sector, and whether or not it is shrinking. If the State is too involved in the labour market, this type of worker data will provide guidance on same.

Over the years, successive administrations have acknowledged the need to restructure the CSO to enable it to recapture its position as the premier statistical office in the Caribbean. Several studies, stakeholder consultations, meetings, committees, and reports have culminated in the decision to replace the CSO with the National Statistical Institute (NSITT). It is anticipated that the NSITT will address the legislative, organisational structure, staffing and IT deficiencies of the current organisation. The expectation is that the new organisation will be resilient and dynamic enough to meet the changing needs of data users.

It is imperative that the NSITT Bill that has been languishing before the JSC for over a year be sent back to Parliament. Further, both the current administration and the Opposition need to put their individual agendas aside and put country first to ensure the passage of the NSITT Bill both in the Upper and Lower Houses.

This piece closes by encouraging the Prime Minister, under whose purview the CSO and its restructuring falls, to do the right thing. Data is the key to development. Evidence-based decision-making by the Government and businesspeople cannot be made without data. Investors cannot invest in our economy without data. Rating agencies cannot report on our economy without data. A well-functioning, well-funded, independent statistical office is the hallmark of a forward-thinking economy.

Dr Roger Hosein is a senior lecturer in Economics at the University of the West Indies.