Sugar-cane is produced by farmers, while sugar mills extract sugar from cane. Therefore, to ensure sufficient quantity of sugar, procurement price of cane, also called the mill-gate price, is to ensure that farmers grow sufficient quantity of cane and are motivated to bring it to mills.

An attempt is made here to analyze sugar-cane procurement pricing during the last 29 years and discuss whether farmers can use it as a guide for future planning.

The officials fix the procurement price of cane each year. Since sugar-cane is a three-to-four year crop, i.e., once planted, cane stalks are cut and crushed each year, while the roots are left in the ground which grow and thus the next crop is produced. Due to this process, land remains dedicated for full three to four years; no other crop can be grown on it. Although, once sugar-cane is harvested, wheat can be grown in-between the rows of cane-roots for getting an additional income.

For this reason, cane procurement price should be announced not yearly but for a period of three to four years. Since this is not done, neither farmers nor mills can make any long-term plan. One thing is however known to them through experience that, as a matter of policy, procurement is never lowered. And hence, while deciding whether to grow sugar-cane, a farmer tries to visualise cane prices for the next three to four years.

Such forecasts can be made by studying the past pattern of prices and using it as a guide to estimate future prices. For example, if changes in official prices were simple increases every year, farmer would assume that the same increases would continue. But, if prices have behaved in a complicated manner, this simple method would not be of any help and computer models will have to be used.

Now let us assume that a farmer is a professionally qualified economist. Will he be able to forecast the procurement prices? Let us examine the situation in some detail.

The economist farmer will follow the same procedure that would be used by an illiterate farmer, i.e., study the past pattern, trying to identify the trends and then use it to make an estimate about the future prices. The only difference will be that he will be using a sophisticated.

Such modern methods are used in a number of disciplines to make accurate predictions. The beauty of such methods is that two or more economists using any of these methods independently will arrive at the same answers.

While even the most complex phenomenon can be represented with the help of some mathematical equation, there is one special case where all such equations and models, including the most sophisticated ones fail— that is called a Drunk’s Walk.

The problem is formulated as follows: suppose that a drunk is standing along-side an electric pole and wishes to walk towards his home. He is fully drunk and is unable to identify the direction of his home, but he does make efforts and tries to walk. The question is, “can we predict where he will be after taking his next three steps?”

When you study the drunk’s previous steps you find that there is no well-defined pattern of movement. He takes one step forward, stumbles back or sideways—it may be two steps forward and one back, or one forward and two back, etc. He may move sideways or in any direction at each given step. As such, there is no way to predict where his next three or four steps will fall. How do you make predictions in a situation that is consist of nothing but just random behaviour?

Statisticians have offered a method to make predictions here also, but there is no guarantee that the predicted results will be correct, or that two or more experts will arrive at the same answer. This method is based on the use of random number tables.

A random number table is a list of numbers in random sequence such that nobody can guess which number will follow next. These tables are available in printed form. While computers can generate their own random numbers, no two computers will ever produce the same set of random numbers, and the same computer when used repeatedly will always produce different sets of random numbers. In practical life, such random numbers are used for drawing lottery tickets, etc.

To forecast the next few steps of the drunk, random number table can be used. We first to use random numbers of two digits. Then we decide to adopt a simple principle depending on our whim,mood or fancy which says that if the first digit of the selected random number is even, the drunk moves forward; if it is odd,he moves backwards. The next digit of the random number is then used to decide how many steps in that direction he will take.Or,it can be vice versa i.e. the first number used to forecast the steps moved,while the second number used to indicate the direction.To simplify the problem,a step taken sideways can bre counted as no step forward or backward mvement.

So the first random number is 41, we may say he will move forward by one step.If the next number is 72 we say he will move back two steps.If the following number is 69 we will say he moves forward by 9 steps.If the number is 30,we say no step has been taken at all.In this way,we select three or four random numbers and estimate where he will be after each step.

The problem here is that two or more professionals using this approach, but working independently, will come up with totally different answers because they may use different assumptions, and may use different random number tables. In other words, it is impossible for even professional economists, using sophisticated software and computers to predict and agree where the drunk will be when he takes the next step. Nothing surprising.

Now let us see how our economist farmer uses any of the above methods to predict the future procurement prices.

Column 2 of the following table gives the procurement price of cane since 1980-81, while col. 3 shows the yearly changes.

Table - Procurement price of sugar-cane 1980/81 - 2002/03

Given these procurement prices for the last over 20 years, the economist farmer would like to forecast what the procurement prices will be during 2003-04, 2004-05 and 2005-06.

Let us try to find out the logic in the prices. Col. 3 shows the yearly changes. We see that sometimes the price remains constant for three or even five years at a stretch, sometimes it moves by less than a rupee, and once it jumped by over ten rupees in one move.

We can say that in the first five attempts, there was no forward movement at all. In the sixth attempt, two forward steps were taken. In the 7th attempt no forward step was taken. In the 8th three-quarters of a step was taken. And so on.

In the 17th we see a unique occurrence as ten and three quarters steps are taken in one attempt - “a giant leap forward.” But then in the subsequent three steps there is again no forward movement at all - perhaps he got too tired after his previous stressful ‘giant leap’. And then in the 22nd step he moved only one step forward - again showing loss of energy.

So we see that the movement of cane prices during the last 20 years shows no pattern, no logic. So, what will the prices be in the 23rd, 24th and 25th years?

If we are forced to provide some answer, any answer, we will have to go to the random number table and select three random numbers. The numbers we get, the assumptions we make, etc., depend upon our sweet wish and whims. And we use them to estimate the future cane prices. The reader may challenge this approach and say that there is no backward movement in this case.

We will get backward movement when we adjust the above prices for inflation. Thus, when we take into account the effect of inflation, using 1980-81 prices as the base, the procurement prices actually fell during the following five years, i.e., there was backward movement. Similarly, whenever the price increase was, or is, less than the rate of inflation, the procurement price in real terms actually moved backwards. In this way, we get both forward and backward movements. To keep the example simple, we have not shown the prices in real terms, but the logic is clear.

The above analysis shows that neither the illiterate farmer nor the economist farmer can forecast what the prices will be in the next three to four years. How would they then decide how much land to allocate to sugar-cane to maximize their gross revenue and net income?

Now, if a hundred thousand farmers use a hundred thousand different random number tables, with a hundred thousand different assumptions, where will the sugar economy end up?

Exactly where the sugar economy of Pakistan is today: offering all types of concessions to encourage import of huge quantities of sugar one year, and then spending so much money subsidizing exports of the same sugar that it causes a dent in the yearly budget! Sugar mills refusing to buy sugar-cane from the farmers, and farmers refusing to sell their standing crop to the mills and allowing it to dry in the sun! Farmers growing huge quantities of sugar-cane, and then, being unable to sell it to the mills, burning it! Subsidizing sugar exports and deciding that cane crushing will be delayed this season.

The above analysis shows the predicament of the farmer and the consequent problems of the sugar industry, and of Pakistan.

We leave it to the reader to decide whether sugar-cane pricing policy being pursued should be classified as rational planning or a .......?