The Hindu Business Line : Controlling growth of disparity

In an FAQ on curbing variance, rich-poor, urban-rural, P. V. INDIRESAN suggests certain rule-based controls instead of the current ad hoc solutions that are neither transparent nor have in-built checks and balances.

What is the Planning Commission Deputy Chairman's own view of the economy?

In a public lecture in Delhi, he said that it is half full and half empty: Half full because the economy is growing rapidly; it is half empty because disparities — inter-State, intra-State; rural-urban, rich-poor, gender — are all increasing.

What is the Planning Commission doing about it?

Apparently it is still looking for solutions.

What is the problem?

The market is a powerful tool to maximise growth; economists have not as yet found a similar tool for controlling disparity.

How does the market maximise growth?

Suppose there are 20 suppliers all supplying identical goods; and the prices they charge are fully known. Suppose, further, all suppliers are equally well accessible. Then, customers will obviously go to the cheapest seller compelling all others to follow suit. In that manner, prices become lowest. When prices are the minimum, sales become the maximum possible. Then, real growth becomes maximum.

Are there no similar techniques for checking disparities?

In engineering, disparities are known as variance. There are standard techniques for minimising variance. Voltage stabiliser is a simple example. Guidance systems for inter-planetary probes operate on the same principle even though they are much more complex.

In nature, the way the human body temperature is kept virtually constant whether the person is rich or poor, white or black, residing in the tropics or in the Arctics, is a classic example how differences are all but eliminated.

How does variance minimisation work?

It is done in three steps: One, the average of the output over all elements is measured. Two, the difference in the output of each element is compared with that average and the difference is estimated. Three, a force is applied in proportion to that difference.

Can we apply a similar technique for minimising, say, inter-regional disparities?

Certainly, provided we know what is the output we want to measure. In the case of economic growth, growth in wages would probably be a better measure than growth in the State Domestic Product (SDP). SDP includes good, bad and anti-bad.

It includes the value of roads built (good), proceedings of crime (bad) and the increase in police force (anti-bad). That is confusing, as we would like to concentrate on good things only.

Growth in wages is less problematic. Real-estate prices too can be used as output indicator, or a combination of both. The choice of output measure is subjective; no fool-proof choice exists.

Suppose we opt for wage growth, how can the regional variance in the growth of wages be minimised?

As a first step, the average growth rate of wages across the country is measured. Then, the growth rate of the particular region (State or even district) is measured.

If the growth is faster than the national average, a disincentive is introduced. If the growth rate is less than the national average, proportionate incentives are provided.

What incentives, disincentives can be used in this case?

The Prime Lending Rate can be adjusted in proportion to the difference in the regional growth rate and the national average. In that case, private investment will tend to shift to low growth areas and away from high growth areas. Thus, the disparity between the two will decrease.

If growth rate is the selected measure, over what period should the growth be taken?

Year-to-year is the obvious choice. Introducing, in addition, growth over longer periods, say, from over one decade (or even over one generation) has important theoretical advantages. A good system will give weight both for year-to-year variation and for decadal growth too.

More sophisticated systems will have a third parameter — actual wages in addition to growth rates. How the weights should be distributed among the three parameters is a matter of judgement for which deeper analysis is required.

In what way does this method differ from the way economy is normally managed?

Manipulating interest rates is one standard way of fine-tuning the economy. However, in that case, no distinction is made between fast growing Bangalore and stagnant Koraput district.

Currently the Reserve Bank of India treats fast growing and stagnant areas alike. That helps overall to control inflation but is of little use in controlling disparity. In fact, a general across-the-board control often worsens disparities: Fast growing areas adjust to changes better than stagnant ones do.

If the solution is so obvious why is it that it is not used?

Some corrections are applied even now: Stagnant areas get subsidies; fast growing areas become more expensive. At the same time, the rich have more clout. Delhi can demand and get Rs 20,000 crore for the luxury of a metro; Koraput does not get anything even for necessities such as potable water, schools, roads, or healthcare. Good infrastructure attracts private investment; poor ones deter investors. Local politics is another factor.

Mostly, ad hoc solutions are used to check disparities. They are not transparent; they do not have in-built checks and balances. For both reasons, case-by-case solutions are not effective — they should be used in the rarest of rare cases. Rule-based controls on the lines suggested can be made powerful enough to minimise disparities as much as desired.

What are these rules?

One, there will be no single national rate of interest or tax; both will be adjusted for disparity. Two, the disparity is measured with respect to the national average. Three, the correcting force is proportional to that disparity. That is, it should be an incentive or disincentive as the case demands. In contrast income-tax/interest changes are one sided. In addition, the incentive/disincentive should be transparent, automatic. Further, it should be frequently updated. Mathematical rules are available to guide all such decisions.

Can this principle be applied to check the increase in rich-poor disparity?

Certainly. In that case, in addition to income-tax, a new levy on the increase in incomes over the previous year should be imposed. Such a tax (positive for excessive increase and negative for less than average) will have salutary addressing effect on growth of rich-poor disparity.

Why not control absolute disparity and not merely growth in disparity?

Controlling absolute disparities is difficult; absolute equality is not even desirable.

(To be continued)

(The author is a former Director of IIT Madras. Response may be sent to: indiresan@gmail.com)

(This is 183rd in the Vision 2020 series. The previous article was published on August 21.)