It was around nine years ago when a book, written by Morten Jerven, then an assistant professor at the school for international studies at Simon Fraser University in Vancouver, stirred a debate on data accuracy of Africa's economy and development. Titled, 'Poor Numbers: How We Are Misled By African Development Statistics and What To Do About It', the book unveiled a 'huge discrepancies and alarming gaps' in African figures. Javan's work also found that just three out of 23 national statistics offices had believed their own calculations of gross domestic product (GDP) covered the whole economy and 18 thought they were underestimated. The book further found that in 2011, only 17 out of 47 selected African countries had prepared their own new GDP estimates and 10 had a 'base year' that was within the previous decade. Another finding was more than half of the rankings of African economies up to 2009 was pure guess estimate.
Though Jeven's study focused on Africa, the problems identified in the work are still valid for many other developing countries in Asia. Many of the African countries are yet to get rid of the doubtful data as overestimation and underestimation are a regular phenomenon there. Some countries in Asia are also facing the problem of update and authentic data despite the fact that a number of measures have been undertaken to improve quality of data, especially, the macroeconomic indicators.
Quality of economic data largely depends on the data collection and compilation process. It needs utmost care to collect and compile the raw data at the ground level. Next step is compilation where input of raw data is critical. Wrong input may lead to distorted or inadequate final data. Staffs in national statistical bodies as well as other national institutions need to be trained properly so that they can collect and compile data accurately. In the collection stage, the staffs have to focus on the reality of the situation and collect what they find. In the compilation phase, it is also crucial to input the data collected from the ground level. The collector and compiler of the data have to ensure the authenticity of the data without thinking about the output. In the data processing phase, it is necessary to cross-check the inputs randomly and go back to compilers and even collectors if necessary. The cross-checking helps to reduce the possibility of errors and omissions.
All this work need meticulous effort. With the modernisation of data collection and compilation procedures, it has now become easier to monitor the phases. For instance, the Bangladesh Bureau of Statistics (BBS) has GPS-enabled tracking system of the field staffs. When they work in the field to collect raw data using tablet or other electronic device, the device can be tracked from headquarter to find out at which spots or places the staffs are currently working. This kind of technology helps to ensure data accuracy and transparency. Again, it is also necessary to deploy the right persons at the right places. Providing national statistics is primarily the preserve of the statisticians. They know better than anyone about the statistical method. By appointing more statisticians, national statistical office can improve the quality of data. Also, regular training of non-statistician officials and staffs is necessary.
Nevertheless, it is ultimately the decision of the policymakers that matters in supplying quality data in a transparent manner. What the policymakers think and how they want to generate quality national statistics and relevant data are the two core issues in this regard. No doubt, sometime policymakers feel uncomfortable about national statistics that portrays bad performance of the economy or any of its sectors. In Bangladesh, there are instances of the release of national statistics getting delayed mainly due to not getting clearance from the authorities concerned. For instance, BBS released the final GDP estimation for FY20 almost a year later, in the August this year. It also made the provisional GDP estimate for FY21 public which is usually released before the end of a fiscal year and quoted in the budget speech. This kind of delay also raises question about data authenticity and even sometime gives rise to e speculation over data manipulation.
Not only the national statistical agency, some other important national institutions have also some problem with data and statistics. An English daily published a report last week on the central bank's overestimation of the country's foreign exchange reserve. According to the report, the International Monetary Fund (IMF) found that Bangladesh Bank has overstated its foreign exchange reserves by $7.2 billion 'through inclusion of non-reserve assets underestimating related risks.' IMF has a clear guideline on foreign exchange reserve for all the member countries. Generally, international reserves comprise foreign currencies, other assets denominated in foreign currencies, gold reserves, special drawing rights (SDRs) and IMF reserve positions.
Though any response from the central bank in this regard is yet to be made available, this kind of overestimation is disturbing and misleading. The overestimation gives a wrong impression about the strength of the foreign exchange reserve. For instance, foreign exchange reserve worth US$ 48 billion (at the end of August) was considered enough to cover seven months' import of goods and services, according to the July-August balance of payments (BoP) table prepared by Bangladesh Bank. By deducting $7.0 billion from the 'overestimated' amount, actual reserve should be $41 billion which is equivalent to payments for around six months' import of goods and services. Definitely, the overestimation of reserve makes the monetary authority more comfortable about foreign exchange management, but the downside is it exposes the economy to the risk of absorbing any undue pressure. One may recall the formal celebration event, organised by the former governor almost a decade ago, to mark the attaining of foreign reserve worth US$10 billion.
The core issue is that be it the national statistical agency or the central bank, an increased focus on strengthened national statistical capacity and data management is a must. It is difficult to ignore any number related to national economy and it is more difficult to overcome any damage due to a wrong number.