Remember elementary Maths? We started with an equation that went:
CP+P = SP

CP – Cost Price
P – Profit
SP - Sales Price.

Now, assume a section where "SP is less than CP" i.e., cost of production is higher than the selling price . Assume further that this "SP less than CP" is not because of market dynamics, but because of government regulations.

That means, that the equation now looks like this:
CP + -(P) = SP
-(P) obviously refers to the continual, unending loss being made by the sector.

However, to correct this situation, the government does this:
CP = CP1 + S.
CP1 - Actual cost borne by the producer.
S – Subsidies.
i.e., we will provide part of your expenditure, such that you do not suffer a loss.
In this, the assumption is that, at least, S is provided to the extent where CP1 becomes less than SP. (i.e., the producer's cost is less than the controlled selling price)

Flawed Assumption #1: "CP1 less than SP. "

Suppose further that the amount of S as per the government’s books is XXXXX, but the amount that actually gets disbursed as S is XXX. i.e.,

CP = CP1+ XXXXX – in the books of the state.

CP = CP1 + XXX for the producers.

Axiom – CP1 must rise to the extent that S is reduced, because the cost of production is not impacted by the availability of subsidies. If the state does not bear the cost, the producer must.

The government fixes SP on the basis of equation 1 above. (CP = CP1 + XXXXX) but the farmer has to bear CP to a much higher extent.

Therefore, no matter what the quantum of S, because of the inherent flaw in the process of calculation, the government will never find out the exact S required, and will also, never be able to fix a SP such that the producer suffers no loss.

That, imho, ladies and gentlemen, is one of the main problems with the Indian agriculture sector.

And if that stupidly drafted post made no sense, all one is trying to say is that control prices in agriculture go against a very basic principle of economics, where a producer must get returns for his efforts.