The Consumer Price Index (CPI) has shown remarkable stability in Pakistan in recent years; in fact, the rate of increase has tended to decline somewhat.
The rate of increase in the Index, which stood at 5.7 during 1998-99, was down to 3.3 per cent in 2002-3. For the entire period, the increase was of 14.9 with an annual average of 3.7 per cent.
This has been achieved, despite a phenomenal rate of monetary expansion, a lackluster rate of growth of the economy, higher import prices, increased burden of indirect taxes and a high rate of population growth. This flies in the face of traditional economic theory and a monetarist should be kicking himself to see such a phenomenon. Table-1:
The sharp disparity in the variables and its unusual limited impact on consumer prices, which is the nature of a paradox, needs to be explained. Partly the explanation lies in the nature of monetary expansion. The bulk of the increase in money supply has been due to the foreign sector, which accounted for as much as 75.8 per cent of the increase. This, in turn, has been primarily the result of spectacular increase after 9/11 in workers’ remittances from abroad. They shot up from $1.1 billion in 1998-99 to $4.2 billion in 2002-03.
The basic difference in origin of the increase and the nature of remittances was significant for the price situation in the country. Previously the bulk of the remittances originated in the Middle East and were mostly meant for family support in rural areas. This affected the price of commodities and services used by the common man and reflected in the CPI. The recent increase has been recorded in the USA, UK, EU and other developed countries and has been prompted by the radically changed environment for Muslims in these countries and crack down on money laundering. In case of the USA alone, the remittances jumped from $82 million to $1,238 million during this period. The remittances from UK rose from $74 million to $274 million over this period. The increase, therefore, has been mainly in the nature of reverse flight of capital into Pakistan. Money has ended up with the affluent in urban areas, from where they had originated, and they are interested in assets, rather than commodities and services of daily use, and they do not form part of CPI.
This is evident from the boom in the stock market, increase in gold prices and in real estate, a process requiring a longer time to work through the economy. Even so, increase in liquidity is also reflected in currency in circulation, which increased by no less than 71.9 per cent during this period, with an average annual increase of 18.0 per cent or more than 4 times the growth rate. This was expected to affect the CPI.
The main explanation of remarkable stability in the CPI is provided by its inherent methodological rigidity. As pointed out in an earlier article (Dawn, July 7, 2003), the methodology used for compiling the CPI is such that it does not reflect the situation on ground. In that article the main focus was the weights assigned to various components of the consumer basket. Equally important are the prices that go into the CPI. This is an attempt to highlight the quality of data and the manner it is used.
Retail prices are collected for commodities and services on a monthly basis from 35 urban centres in the country. Their average goes into the Index. The market basket consists of 10 broad groups and the number of items included therein are: food\beverages and tobacco (124), apparel\textile and footwear (42), house rent (1), fuel and lighting, household (15), furniture\equipment (44), transport\communication (43), recreation\entertainment (18), education (22), cleaning\laundry and personal appearance (36), and medicines (29).
The centres covered are: Lahore, Faisalabad, Rawalpindi, Multan, Gujranwala, Sialkot, Sargodha, Islamabad, Jhang, Bahawalpur, Bahawalnagar, Okara, Jhelum, D.G. Khan, Mianwali, Attock, Samundri, Vehari, Karachi, Hyderabad, Sukkur, Nawabshah, Larkana, Mirpurkhas, Shahdadpur, Kunri, Peshawar, Mardan, Abbottabad, D.I. Khan, Bannu, Quetta, Khuzdar, Turbat, and Loralai.
A simple average of the prices prevailing at these centres is taken. The basic limitation of simple average is that a few large variations can swing the overall average for the sample depicting a very distorted, if not misleading, picture.
The range of variation in inter-city prices and the way these are collated for inclusion in CPI makes an interesting study. It would be very revealing if the information is presented in a matric form. However, this is not possible in an article due to the space constraint. Nevertheless, a sample of a few basic food commodities is presented (Table 2) giving their maximum and minimum prices indicating the places, with those in Karachi and the average for 35 cities in June 2003. They are compared with those a year earlier and in 1999, and have been obtained from the various issues of FBS Statistical Bulletin.
With a wide variation in inter-city prices, which in some cases is 1 to 4, taking their simple average is unrealistic, to say the least. To quote a specific example how just one quotation can distort the whole picture, we may take tuition fee for government Eng. College Ist Year fee in June, 2003. For Faisalabad and Samundri it has been shown as Rs3,000 while it was Rs125 at Peshawar, Rs.250 at Nawabshah, Rs.293 each at Quetta and Khuzdar, Rs407 at Sukkur and ranged between Rs500 and Rs600 at other cities. At Karachi it was Rs500. This gave the average fee of Rs799.50.
For such situations statistics has a concept of “median” which indicates the level around which the most prices cluster. This is the relevant measure for Pakistan. Combining the weights assigned to various components of the consumer basket, which shows how much is spent and the prices of commodities and services, which indicate what they actually get, one gets a very preposterous picture. In June 2003, an average family of six persons in the lowest income bracket of Rs3,000 per month could buy in a month 16 kilos of wheat flour, 0.5 kilo of mutton or 1.3 kilos of beef or live chicken, 1.6 kilos of vegetable ghee (loose), 3 kilos of sugar, 11.3 litres of milk, 1.5 kilos of Basmati (Av. Q) or 3.3 kilos of Irri rice.
It should be obvious from the preceding analysis that not enough care is taken in collection of price data. For the tuition fee quoted above, one wonders if there is an engineering college in Samundri and Khuzdar. There are far more interesting, rather amusing, other cases. For instance, hair cut for men cost in 2003 Rs37.50 in Turbat as against Rs28.87 in Karachi. Curd is always Rs3-4 more expensive than fresh milk but in Turbat fresh milk cost Rs20 while curd was twice as much at Rs40, the highest in the country. In Karachi the respective prices were Rs23.00 and Rs32.00.
It would be stating the obvious that the present CPI does not reflect the ground realities. The simple average price for all centres is irrelevant to big cities in the country, where the bulk of the urban population resides. This becomes quite obvious when the average price of a commodity is compared with that in Karachi. It may be observed from the matrix that not only the prices in Karachi have been much higher than the average taken, but also increased faster.
The urgent need to overhaul the CPI drastically, to make it realistic and scientifically correct cannot be over emphasized. Prices should be weighted by population of the city and cover all items. The present CPI is based on tuition fees charged by government institutions as if private educational institutions do not exist in the country. Elite institutions have a mushroom growth in urban areas, charging fabulous fees and private tuition for their students has become a must. Then there is new phenomenon of school transport costing a tidy sum. The cost of laboratory tests, so common these days, is a conspicuous omission in the medical group of the Index. Moreover, it is perhaps believed that an MBBS doctor would do and there is no need to go to a specialist doctor. The house rent does not indicate the trend of the amount of rent paid, but the cost of construction. It is not clear how this can serve the purpose when the two are inversely related. These are a few glaring cases.