After the dark age of the 1990’s Romania’s economy started to stabilize and with it the inflation began to slowly but surely decrease down to below 5% in 2006. In this context many consumers started to feel that the inflation was actually higher that official statistics reported. Add to this the traditional distrust in central authorities and you have a widespread sentiment among Romanians that the National Bank of Romania (BNR) is lying in its inflation reports. Another more hidden factor can be well summarized by the old Romanian saying that “the Romanian is good at everything”. Many consumers compare their own perceptions that result from their spending habits and observations with the scientifically calculated statistic that comprise the entire economy. But even so, in theory, the average sentiment of the people should confirm the statistic, which it doesn’t. However, there is a scientific explanation for the higher perceived inflation rate.

The main problem is the weight. In statistics, the weight of a price change is the actual percentage of the basket of goods that is represented by that item. This can be roughly translated in what percentage of its income the average Ion spends on an item. But to the consumer’s perception, what is of more importance is not how much money is spent on an item, but how often it is bought. For example, suppose a basket composed only of food and communication services. According to Eurostat data on Romania, approximately 40% of the basket’s value is represented by the food and around 4% by the communication services. This would correspond to a weight of 90% for the food and 10% for the communication services in my example. Now let’s say that during a month the food prices go up 1% on average and the communication services prices go down 0,7%. (I am sure you find this assumption not far from the Romanian reality of the past few years) Statistically, this would lead to an increase in the consumer price index of 0,83% (1% ∙ 90% – 0,7% ∙ 10%). But to Ion the picture looks a bit different: while he only pays one phone bill and one cable bill, he buys, say, 200 items of food per month. Because he notices price changes every time he pays for something, the weights of his perceived price index will be the frequencies of the purchases and not the actual sums. Thus, the perceived inflation according to this rule would be 0,983% (1% ∙ 200/202 – 0,7% ∙ 2/202) which means that Ion hardly even notices the drop in the prices of communication related services.

Another problem with the perceptions is that the impact any price rise has on the overall sentiment of the average consumer is higher than the impact of a similar price drop. In other words, if Ion wants to buy something and sees that it now costs 10% more than it used to he will be much more upset than he will be happy if it would cost 10% less than it used to. Something like this is not easily quantified, but professors and researchers like Hans Wolfgang Brachinger from the University of Fribourg (Switzerland) have done extensive work on the matter, intrigued by the apparent price level boom after the Euro changeover in countries such as Germany, Austria or Italy.

For details and available research concerning the differences between perceived inflation and the CPI/HCPI see

Alin Beg