Inflation rate is one of the most important economic indicators for both developed and developing countries.
Every country spends a tremendous amount of time to institute policies, both fiscal and monetary to ensure that inflation rate is kept under manageable targets.
This is to ensure that economic growth and development are not removed by uncontrolled inflation.
The Ghana Statistical Service (GSS) is the sole government agency responsible for estimating and publishing inflation data in Ghana. Other government institutions such as the Bank of Ghana, however, collects a limited amount of inflation data for its own use.
Inflation data published by GSS has come under public scrutiny in recent times. There is a perception in the country that inflation data published by the GSS is inaccurate. This paper attempts to highlight the possible causes of this perception from economic and statistical perspectives. The paper suggests the various methodologies for estimating inflation rate and how the GSS could improve this important estimate.
Estimating inflation from Consumer Price Index
The most popular method for estimating inflation rate is through the Consumer Price Index (CPI). The CPI uses a fixed basket of goods from a base year. The quantities of goods and services consumed remains the same from year to year in the eyes of the CPI, whereas the prices of goods and services in the basket change.
This type of index, where the basket of goods is fixed, is called a Laspeyres index. To take care of changes in consumers’ expenditures and consumption patterns, weights that are attached to goods and services in the basket are periodically revised through the conduct of consumer expenditure surveys.
Problems associated with CPI inflation
One of the factors contributing to the inflation controversy is that the statistical methodology for calculating the CPI (which is the measure through which inflation is most commonly derived from) is complex and not easily comprehensible at least by an ordinary person.
In addition, some economists believe that CPI, in particular, overstates the true rate of inflation, because they do not take into account improvements in the quality of goods and services which may offset, in some cases, some of the rise in prices. Furthermore, it may understate the true inflation, due to inadequate allowances for changes in the cost of housing and other hard-to-state prices.
From the statistical perspective, estimating inflation from the CPI is deficient in the sense that the variance of the point estimate is unknown. Consequently, scientists have no way of knowing how precisely inflation is measured.
Given a level of inflation, statisticians are interested in constructing a confidence interval of this point estimate that would contain the true but unknown rate of inflation at a given level of probability. This could only be possible, if we could calculate the variance of the estimate. The variance, if provided ,will be the scientific basis for evaluating the point estimate.
Estimating inflation from GDP deflators
The GDP deflator, on the other hand, uses a flexible basket of goods that depends on the quantities of goods and services produced within a given year. The prices of the goods are fixed. This type of index, where the basket of goods is flexible, is called a Paasche Index. Whereas either the Consumer Prices or GDP deflator indices could be used for the calculation of inflation, studies have shown that neither is perfect.
Estimating inflation based on economic theory
Economists have ways of estimating inflationary trends based on economic theory. One of such common models is the Phillips Curve.
The model relates the unemployment rate or some other measure of aggregate economic activity to a measure inflation rate. Modern specifications of the Phillip’s curve relate unemployment rate, or some other measure of aggregate economic activity to future changes in the rate of inflation.
The idea is that when unemployment is below a baseline rate, inflation tends to rise over time and when unemployment is above this rate, inflation tends to fall. This baseline unemployment rate is known as the Non-accelerating Inflation rate of Unemployment (NAIRU).
This econometric modelling technique of NAIRU Phillip’s curve is widely used to produce inflation forecast to the extent that some economist consider them more accurate than the forecast made by other methods.
Another example of model-based inflation estimation is derived from the quantity theory of money. The model uses monetary aggregates to predict inflation.
How inflation is estimated in Ghana
In Ghana, inflation is estimated from the consumer price index. The index is calculated from a fixed basket of goods and services consumed by the average Ghanaian which is periodically revised using data from the Ghana Living Standards survey, a household expenditure survey.
Economic and Statistical Evidence about Inflation
Like many economic indicators, there exists a relationship between inflation and other economic indicators. Inflation estimates can be arrived at using various procedures i.e. from consumer price index, GDP deflators and based on economic theory. Estimating inflation from these sources will definitely be different in absolute terms.
However, they are expected to move in the same direction over time. Furthermore, irrespective of what source that inflation rate is calculated from, it is usually fairly similar in value. In theory, there is a significant difference between the abilities of each method to capture consumer's consumption choices when a change in price occurs, thus providing the basis for controversies.
Using the Consumer Price Index to estimate inflation should always be accompanied by their standard errors of the estimates. Statistically, the standard errors give credence to the veracity or otherwise of the estimates.
It is not enough for officials of the Ghana Statistical Service to make statements such as ‘the Statistical Service does not manufacture inflation figures’ without coming out to prove to Ghanaians that inflation estimates published by the service are indeed accurate.
I expect the Ghana Statistical Service to first re-evaluate the sampling procedures, data collection procedures etc. with the view to reducing sampling and non-sampling errors.
This aspect could be done by the service using local expertise or utilising knowledgeable consultants to assist them. This is being done elsewhere which suggests that there are best practices to be followed by the Statistical Service and it does not need to re-invent the wheel. In this respect, the GSS should, as a matter of procedure, always publish the sampling errors associated with the inflation estimates.
Finally, it will not be fair to challenge inflation estimates produced by the GSS without providing the basis for such a challenge. Comments such as: ‘go to the market and find things for yourself’ are too naïve and not acceptable in the scientific or statistical ‘world’. Critics should provide evidence and state which method they are using to estimate inflation before challenging the status quo.
The writer is an economist/statistician.
Email:
