The taxman generally relies on the income-tax returns filed by assessees. Insurers rely on the declarations by the insured, and `planning bureaus rely on performance reports from productive enterprises under their direction'.

In these examples, "the reporting agents have an incentive to misrepresent because transfers between the parties depend on their reports," writes Dilip Mookherjee in a chapter titled `Optimal auditing, insurance, and redistribution,' in Market Institutions, Governance, and Development, from Oxford University Press (www.oup.com).

The collection of essays in the book trace underdevelopment to `deficiencies in underlying institutions' and also `contract enforcement mechanisms, information systems, inequality of land and education, and limited government accountability'.

Part two of the book is about `problems in governance in the context of tax evasion, bureaucratic corruption, fiscal decentralisation, and delivery of public services'.

The `auditing' essay, we started off with, seeks to find `general conditions under which random audits are optimal'. All audits must be random in any optimal contract, is a key result of Mookherjee's study. "The intuition for this is that as long as there are non-trivial penalties for those discovered to have reported falsely, an audit probability equal to one ensures that the agent will strictly prefer not to lie," reasons the author.

Another chapter is provocatively titled, `Corruptible law enforcers: How should they be compensated?' It begins with the quote of `anonymous economics professor', thus: "Only after leaving New York did I learn that a municipal sanitation inspector's job was not to collect money from door to door."

Whenever a principal delegates enforcement authority, opportunities for corruption arise, writes Mookherjee. "Sanitation inspectors, auditors, production foremen, and financial regulators all have a discretion to `sell out'." What a company for auditors! A footnote has more damning stuff: In Taiwan, 80 per cent of CPAs (certified public accountants) interviewed by a researcher two decades ago `admitted to having bribed public tax officials'.

What about India? "In India, the Policy Group (1985) estimated that 76 per cent of government tax auditors took bribes, and 68 per cent of taxpayers filing through CAs had given bribes."

Yet another chapter, on `tax administration reform and taxpayer compliance in India', speaks of `evidence consistent with the model of taxpayers strategically self-selecting into wards or circles'. Mookherjee suggests that the ward/circle distinction should be replaced with `random assignment rules'. He argues that such a move will remove the strategic under-filing incentive, with beneficial compliance effects.

`Other more sophisticated ambitious reforms' that he discusses are: "Reducing discretion of assessing officers with regard to selection and conduct of audits, reforming performance evaluation and personnel allocation procedures, centralised audit selection procedures based on income disclosures and information generated by third parties."

For accountants who are not averse to economics.