|
Blog: 24 March 2017 (Inflation of the rich and poor in South Africa)
|
Related Topics |
|
In today's blog we take a look at South Africa's inflation rate in more detail. We look at the inflation rate per income group. From the inflation rate of the poorest 10% of households to the inflation rate if the richest 10% of households. And we find that the difference in inflation experienced by the different groups is contributing to the continued inequality experienced in South Africa.
|
Rich vs Middle Income vs Poor
Since Statistics South Africa (Stats SA) breaks inflation down according to 10 different expenditure groups (deciles) we can now get inflation rates for the poor, middle income households and the rich. The deciles are calculated based on household expenditure. Expenditure per household was ranked from lowest to highest. From there the households were grouped into 10 different groups based on where they were ranked. The contributions of the 10 deciles to total expenditure in South Africa is as follows:
|
From the table it is clear that the contribution to overall spending in South Africa is totally dominated by the richest 10% of households, with their spending making up almost half of total spending in South Africa. While the poorest 30% of households spends less than 5% of all money in South Africa. It is no surprise then that critics of South Africa's inflation rate calls the CPI basket a basket of the rich.
So over time what does the inflation rate for the poor vs rich look like? For presentation purposes we will not be plotting the time series of all 10 deciles in the graphic below but rather focus on the poorest 10% (Decile 1), the mid point or for this article our middle class (decile 5) and then the rich which is decile 10. The line graph below shows the year on year inflation rate experienced by the aforementioned three groups from Janaury 2009.
So over time what does the inflation rate for the poor vs rich look like? For presentation purposes we will not be plotting the time series of all 10 deciles in the graphic below but rather focus on the poorest 10% (Decile 1), the mid point or for this article our middle class (decile 5) and then the rich which is decile 10. The line graph below shows the year on year inflation rate experienced by the aforementioned three groups from Janaury 2009.
From the line chart it is clear that the poorest of the poor has experienced a continued higher inflation rate than that of both middle income earners and the rich. With the middle income earner's inflation rate being lower than that of the poor but higher than that of the rich. And lastly the average level of inflation experienced by the richest 10% of households in South Africa is far lower than the average inflation level of the poorest 10% of households. Just perpetuating the already wide inequality gap in South Africa. How you may ask? Well if the price of items the poor by regularly increases at a faster rate than the items the rich buy, the buying power of the poor is decreasing at a faster rate than that of rich and thereby widening the inequality experienced in South Africa. As we will show in the graphic below the average rate of inflation for the poorest 10% of households amounted to 6.4 % a year while for the richest 10% it was a 5.5% a year. A difference of 0.9% per year for 8 years. In total prices of goods and services for the poorest 10% of households increased 7.2% more for the poorest 10% compared to the price increases of the richest 10%.
Average level of inflation from January 2009 to February 2017 for the 10 different deciles are shown below. And it is clear as you move along the chart towards the right (towards the rich households), the average level of inflation experienced drops substantially.
The bar chart also shows the official annual inflation rate over the period January 2009 to February 2017. And this level is very close to that of Decile 9 and Decile 10 (the richest 20% of households). Reinforcing the view that the CPI basket as published by Stats SA is an inflation rate for the rich and not the majority of South Africans. In addition to the inflation of the poor being higher than that of the rich, it tends to be more volatile too (as the goods and services they consume tend to have more severe price movements). To illustrate this we will calculate confidence intervals for the 10 Deciles. Confidence intervals is statistical concept used in which statisticians can calculate a range within which a variable would fall and say that 95% of the time the variable would fall within this range, or 98% of the time it will fall within this range.
We will calculate 95% confidence intervals. Thus 95 out of 100 times the inflation rate per decile will fall within the confidence intervals calculated. Below the upper limit, lower limit and average inflation rate per Decile.
We will calculate 95% confidence intervals. Thus 95 out of 100 times the inflation rate per decile will fall within the confidence intervals calculated. Below the upper limit, lower limit and average inflation rate per Decile.
What is clear from the graph is the fact that the range is wider for the poorer deciles and becomes narrower for the richer deciles. Supporting the view that the rate of inflation experienced by the poor tends to be a lot more volatile than that of the rich. So what if all deciles were given equal weight and the CPI was recalculated? What would the overall level of inflation be at? As the overall level of inflation is skewed towards the rich by the fact that they carry such a large weight in the overall basket. One wonders if government is aware of this and if they adjust social grant payouts at a rate greater than inflation as published (as this clearly shows it needs to be the case as the official inflation rate published does the poor absolutely no favours.
The graphic below takes a look at the published rate of inflation vs our recalculated inflation giving all deciles an equal weight.
The graphic below takes a look at the published rate of inflation vs our recalculated inflation giving all deciles an equal weight.
Equal weighted Deciles. What would official CPI be at?
From the line graph it is clear the recalculated level of inflation (if all deciles were given an equal weight) would be higher than the officially published number. No wonder the masses keep saying the published inflation figure does not reflect their reality. Over the course of the period (Jan 2009 to Feb 2017) the average inflation rate came in at 5.6% while the recalculated inflation rate came in at 6%. That's 0.4% difference each year for 8 years. Imagine the impact on inflation linked salaries, bond repayments etc if this was the officially published number.