DINESH RAMBALLY
AS BRITISH former prime minister Benjamin Disraeli once famously quipped, “There are three kinds of lies: lies, damn lies and statistics. The phrase serves as a warning that we should not be too eager to accept statistics as trustworthy because there is a myriad of ways in which it can be manipulated to bolster weak arguments.
One such glaring recent example of such data-reaching was that employed by Minister of Finance Colm Imbert to show support for his statement that “Trinidad and Tobago is not falling apart.” He used the Central Statistical Office’s (CSO) recent gross domestic product (GDP) data release to show there was economic growth by quoting a $5 billion increase in the country’s figures from the second quarter of 2021 to the third quarter of the same period.
Now what was interesting was that the growth of $5 billion, while more or less accurate (it was actually $5.6 billion according to the figures on the CSO website), was that it represented the increase in current price GDP rather than constant price GDP. This statistic was conveniently used to build his entire narrative of an economy that was seemingly well managed by his political party, the People’s National Movement (PNM).
GDP by definition is the monetary value of all final goods and services produced by an economy over a specific period of time (usually a year or a quarter). While GDP is collected at current or nominal prices, in order to determine if the economy is expanding or contracting (what economists term real GDP or constant price GDP), current price GDP has to be adjusted for price changes.
This is done to ascertain if the increase in value is due to a rise in output (because more is being produced) rather than an increase in prices. As a result, constant price GDP should be used instead as it is adjusted to take into account price changes so that it will be easy to tell if actual output is growing or even shrinking.
As stated by the CSO, constant price GDP increased by 8.9 per cent (a $3 billion increase) from $33.6 billion in the second quarter of 2021 to $36.6 billion in the third quarter of the same year. Again, while this seems feasible especially given the lockdown in the second quarter (which would have suppressed output), the quarterly data to my knowledge was not adjusted for seasonality (predictive fluctuations), rendering the data unreliable.
One problem that economists/statisticians face when analysing data is that because many data series exhibit movements that reoccur every year in the same month or quarter, it becomes difficult to interpret the underlying trend.
For example, retail sales usually rise in December because of the Christmas holiday. To determine the actual trend in sales from November to December in this case, the predictable Christmas run up from November to December has to be removed. Likewise, to understand what the CSO data is saying from the April to June quarter to the July to September quarter, the predictive fluctu
