Chapter 20, in the middle of a passage about non-attachment, stops and gives a household budget.
One's own wealth should be expended in three ways — for dharma, for enhancing the wealth and for objects of pleasure.1
A third to religion. A third reinvested. A third spent on living.
No hedging, no "according to capacity," no alternative for the poor. Three parts, named, in order.
It is the most ordinary sentence in the chapter and one of the more consequential, because it is the point at which a devotional text becomes a personal finance manual and does not appear to notice the transition.
Each third gets a specification, and they are not equally detailed.
Dharma is itemized: "For dharma, one should perform nitya karma, naimittika karma and kamya karma."2 The daily obligatory rites, the periodic occasional ones, and the desire-driven optional ones — the standard three-fold classification of rite, given elsewhere in this cluster with its own timing doctrine.3
Enhancement gets one clause: "The second part must be spent in the beneficial act of enhancement of wealth."4 Reinvestment, unelaborated.
Enjoyment gets one clause: "The final part can be spent on the enjoyment of objects of pleasure."5
Note the modal verbs. Dharma should be performed. Enhancement must be spent. Enjoyment can be spent.
So the three thirds are not equally binding, and the loosest is the one the same chapter has just made causally necessary — enjoyment being the stage through which properly-earned wealth produces non-attachment.6
Can be is an odd verb for a required step.
Then a second layer, and this one is occupational rather than proportional.
One tenth of what has been earned through agriculture must be spent on cleansing the sin. The remnant can be spent on dharma and the other things. Otherwise, a person will go to Rourava.7
A discriminating person who earns his subsistence through usury must set aside one-sixth for donations. Pure and excellent brahmanas who receive, must give away one-fourth as donations. If wealth is received unexpectedly, an excellent brahmana will give away half.8
Four rates:
| Source of income | Rate |
|---|---|
| Agriculture | 1/10, before anything else |
| Usury | 1/6 |
| Brahmana receipts | 1/4 |
| Windfall | 1/2 |
And Debroy supplies the reasoning behind the first, which the text withholds: "Presumably because agriculture has caused violence to living beings."9
That is a guess, and he marks it as one. But it is the right guess, and it makes the agricultural tenth a different thing from the other three. It is not a donation rate. It is compensation for damage done in the earning — which is why it comes off the top, before the thirds, and why failing to pay it sends you to a hell rather than merely reducing your merit.
The windfall rate is the interesting one, and it is the steepest.
If wealth is received unexpectedly, an excellent brahmana will give away half.10
Half. Against a quarter for ordinary receipts and a sixth for money-lending.
Why would unearned money carry the highest rate?
The chapter's own logic supplies the answer, and it is the provenance argument running one chapter earlier: artha earned through dharma produces non-attachment; artha earned otherwise produces attachment.11
A windfall has no provenance. It was not earned through dharma or through anything else — it arrived. So it cannot do the work the chain requires of wealth, and holding it produces the bad outcome rather than the good one.
Giving away half is therefore not generosity. It is decontamination, and the rate scales with how unearned the money is: a sixth for interest (earned, but passively), a quarter for gifts received (unearned but expected), a half for pure windfall.
And the sequence completes with the harshest clause in the passage:
Everything received from a wicked person must be flung away into the ocean of charity.12
Not a fraction. All of it. Money from a bad source is not taxed; it is disposed of, and the phrase for disposal is the ocean of charity — which makes charity, here, explicitly a place to put things you cannot keep.
Harvest is in and you sit down with it, as you do every year, and it takes most of an afternoon.
The tenth comes off first. Not a donation — you have long since stopped thinking of it as one. It is what the land costs, in the way the seed is what the crop costs. You have never once considered skipping it and you know men who have.
What is left, you cut in three.
The dharma third is the easiest to spend and the hardest to spend well. Nitya you already do. Naimittika is the monthly and annual business, which is largely food for other people. Kamya is where a man's actual wants become visible in his accounts, and yours are visible, and you would rather they were not.
The enhancement third goes back into the land. This is the part nobody preaches about and it is a third of everything, sitting in a religious text, unapologized for.
And the last third you are supposed to spend on enjoyment, which you find surprisingly difficult, because you were raised to think that part was the leftover and the text says it is a share.
Your brother had a windfall two years ago and gave away half and everyone thought he was mad. He has since said, once, that it stopped being frightening after that.
The budget is a mechanism for making money finish its journey, and the tenth and the half are the two places where the journey has to be corrected before it can continue.
Return to the tenth, because it is doing something the rest of the passage is not.
Every other rate in the list is about the money. The tenth is about the act of getting it.
Ploughing kills. Not intentionally, and not as a matter anyone in the passage argues about — the text does not raise the ethics at all, which is precisely why Debroy had to supply the reasoning. It is treated as a known fact with a known price.
And it sits interestingly against the same chapter's status taxonomy, which demotes a brahmana who ploughs the field himself to shudra brahmana — the second-lowest of the seven grades.13
Put the two together and agriculture in this chapter is: the occupation that requires an ongoing sin-payment and the one that costs a brahmana five grades of standing.
That is a coherent position and an uncomfortable one, since it is also the occupation feeding everybody in the room. The text prices it, penalizes it, and never once acknowledges the dependence.
The disposal clause repays a second look because of what it implies about how money is understood to work here.
Money from a wicked person cannot be purified by a percentage. It has to go entirely.
Which means contamination in this system is not proportional — it is a property of the whole quantity, and no fraction of it can be redeemed by giving the rest away. Compare the windfall, where half is enough. A windfall has no provenance; wicked money has a bad one, and bad is worse than absent.
That ordering is the passage's implicit hierarchy of sources, cleanest to worst:
Six grades, and the chapter states them across two paragraphs without ever presenting them as a scale.
The more distant the money from your own righteous effort, the less of it you may keep.
Last, the structural observation, which is what makes this a page rather than a table.
Every rate in this passage has an enforcement clause, and the enforcement is not social.
Fail the agricultural tenth: Rourava. And "if he does not do this, he is evil-minded, and it is true that he will bring about his own destruction."14
The same hell appears twice more in the same chapter — for the receiver who does not cleanse his half-merit debt, and for the forest-dweller who lapses in his japa.15
So chapter 20's financial rules are backed by exactly the same sanction as its ritual ones, in the same words, in the same paragraph run.
Which tells you that the compiler drew no line between the two. A man's accounts and a man's japa are the same kind of obligation, failing either has the same consequence, and both are self-assessed with no auditor.
This is a tax code whose only enforcement mechanism is the taxpayer's own belief. Which is, in fairness, a considerably cheaper system to run than the alternative — and it explains why the rates are stated so precisely for a revenue nobody was collecting.
Strongest support is the passage's own completeness: three thirds, four occupational rates, a disposal rule, and a stated hell, all within one paragraph run.
Tension one. Enjoyment is a required stage in the chapter's non-attachment chain and gets the weakest modal verb in the budget — can be spent against dharma's should and enhancement's must.
Tension two. A third of income is allocated to growing wealth, unapologized for, in a passage about renunciation.
Tension three. Agriculture carries an ongoing sin-payment and demotes a brahmana five grades, in a text whose audience it feeds. The dependence is never acknowledged.
Tension four. Contamination is proportional for windfalls and total for wicked money. No principle is given for why one is dilutable and the other is not.
Open question. Does the tenth on agriculture come off before the thirds, or is it the dharma third under a different description? "The remnant can be spent on dharma and the other things" suggests before — but if so, a farmer's dharma spending is a tenth plus a third of the remainder, which is a heavier burden than any other trade carries, and the chapter does not remark on it. Debroy translates without resolving the sequencing.
The vault's practitioner Śaiva corpus has no economics at all, and its absence is the point.
Worshipping the Jīva as Śiva makes work itself the worship — the labour is the offering, and no accounting attaches.
Chapter 20 has an accounting for everything, down to a sixth on interest income.
That is a genuine difference in what a religious life is understood to consist of. One says: do your work as worship. The other says: here are the four rates and the hell for missing them.
Neither is a softening of the other. They are different technologies, and the vault should hold both rather than letting the modern one stand as the tradition's settled position — because the primary text is the one with the numbers, and the numbers ran for a thousand years.
Dharma, Artha, Bhoga, Vairagya is the doctrine this budget implements, and the two should be read together: the chain says clean wealth spent on enjoyment produces detachment, and the budget says how much.
Minimum Viable Overhead (business). The operator material treats a fixed allocation — a standing percentage committed before discretionary spending begins — as the mechanism that makes long-horizon behaviour survive short-horizon pressure, and finds that the automaticity matters more than the rate.
Chapter 20's thirds are exactly that structure: a pre-commitment applied at the moment income arrives, before any decision about what is affordable.
What the pair produces: a fixed share allocated at the point of receipt is not a budget, it is a device for removing a recurring decision — and the reason both a medieval Purana and a modern operator arrive at thirds and tenths is that the alternative, deciding each time, reliably fails under pressure. Which reframes the passage: the precision of the rates is not the interesting part and probably never was. The interesting part is that the allocation happens before the money becomes yours in any felt sense, which is the only point at which giving a third away is psychologically cheap.
Mental Accounting (psychology). The behavioral-economics work shows that people spend the same money differently depending on which mental account it lands in — and specifically that windfall money is spent more freely and more regrettably than earned money, because it never enters the account that carries a sense of cost.
Chapter 20 taxes windfalls at half, the highest rate in the passage.
The insight neither gives alone: the tradition priced the windfall problem correctly a thousand years before it was measured — and it did so from a completely different premise, that unearned wealth cannot produce non-attachment because it has no provenance. Two incompatible theories, one correct prediction. That convergence is worth flagging as a methodological point: a moral system reasoning about where money came from and an empirical one reasoning about how money is mentally filed will identify the same hazard, because provenance is exactly what determines the filing. The Purana's rate is not a guess; it is an accurate reading of a real effect, arrived at through a metaphysics that has nothing to do with it.
Sharpest implication. Chapter 20 hands a household a budget — a third to dharma, a third reinvested, a third to enjoyment — with a separate schedule by trade: a tenth off the top for agriculture (because ploughing kills, as Debroy has to supply), a sixth on usury, a quarter on receipts, half on any windfall, and total disposal of anything from a wicked person. Read as a scale it runs cleanest to worst by distance from your own righteous effort, and the rates rise as the distance does. Every one of them is enforced by Rourava, the same hell the chapter uses for lapsed japa — so a man's accounts and a man's practice are the same kind of obligation, failing either costs the same, and both are self-assessed with no auditor anywhere.
Generative questions.