How India's Macro Institutions Fit Together

India's macro arrangements sort into three functions rather than a list of names: measuring, deciding, and publishing the record. The Reserve Bank of India holds the rate lever, the Ministry of Finance holds spending and taxing, and the National Statistical Office compiles the official statistics. The measurer is not the decider, and that separation is what lets anyone else check a decision.

Underneath the names there is a division of labour, and it is the labour rather than the names that is worth learning. Somebody has to produce a description of the economy that is compiled by a stated method rather than felt. Somebody else has to pull a lever that acts on that economy. And whatever either of them does has to end up written down in a place a stranger can walk into and read. Three jobs, then, and India puts them in different places. Keeping the three in three places is what makes any one of them checkable by an outsider.

What are the three functions, and why sort by them rather than list the names?

Begin at kitchen table scale, where the same shape turns up with far less riding on it. One person keeps the accounts book: what came in, what went out, written down the same way every month. Another person decides what the household will spend on next month. And the book stays on the shelf in the front room, where the grandmother, the son who sends money from another city, and anyone else with a stake can open it. Three jobs, three places. Nobody in that household set it up as a system of checks. The separation simply pays. When the book is kept by the person who is not deciding, an argument about last month can be settled by opening the book instead of by raising a voice.

India's macro arrangements have the same three jobs in them. Sorting the bodies by what they do rather than by what they are called gives even an unfamiliar body somewhere to go. A body that compiles a series by a published method is a measurer. A body that can change something about the economy by an act of its own is a decider. And every one of them puts out documents. Those documents are the third job: the record. A name says nothing about which of the three a body does, so learning the list of names instead leaves a reader stuck at the first unfamiliar acronym in a footnote.

The functions also survive change in a way a list does not. Bodies get renamed, merged and given new remits. The three jobs are what an economy needs done rather than what a government happens to have set up, so the three jobs do not go anywhere. So when a familiar body is reorganised, the question to ask is not what it is called now. The question is which of the three functions the body is doing, and whether it is doing more than one of them.

Which body holds which lever?

Two levers act on the economy, and they are not in the same hands. The rate lever, the one that works through the price of money and the credit side, sits with the Reserve Bank of India. The spending lever and the taxing lever, the ones that work through what the government pays out and what it takes in, sit with the Ministry of Finance. The rate lever and the spending and taxing levers are held by two different bodies, and that is the first thing the structure says about India.

Say what kind of body each one is. The kind explains a great deal about how it behaves. The Reserve Bank of India is the central bank, and it is a statutory bodyA body brought into existence by an Act of the legislature, which sets out what it is for and what it may do. The body exists because a law says so, not because a department chose to create it.. A law says the bank exists, and that same law sets out what it is there to do. The Ministry of Finance is a department of the executive governmentThe branch that administers and carries out the law, as distinct from the legislature that makes it and the courts that interpret it., staffed by the civil service and headed by a minister who answers to Parliament. Its spending plans are not its own to enact in the way a rate decision is: money is voted, and the legislature's appropriationThe legislature's formal act of authorising money to be drawn and spent for a stated purpose. Until it happens, a spending plan is a plan. is what turns a plan into permission to pay.

Each lever does a great deal once it is pulled, and that is a long subject of its own. The rate side, the corridor it works inside and how a change travels outward is covered separately, under the policy rate. The spending and taxing side, and what changing either does to the wider economy, is covered separately too. Here the levers only need holders. Knowing that the rate lever and the spending lever are in two different pairs of hands is enough to read almost any macro sentence correctly, and it is exactly what a reader who learned a list of acronyms does not know.

Who does the measuring?

The National Statistical Office sits inside the Ministry of Statistics and Programme Implementation, and it is the body that compiles official statistics. Compiling is a much heavier word than it looks. Compiling means designing a sample surveyA method of learning about a whole population by observing a carefully chosen part of it, so that what the part shows can stand for the whole within a stated margin.. A manageable number of households or units then stands for a very large number of them. The job also means gathering administrative dataRecords thrown off by the ordinary business of government, such as filings, registers and returns. The records were created for another purpose and can then be used as statistical raw material., the records that government already generates by doing its ordinary work. Compiling means sticking to a method that has been published. Only a published method lets one stretch of time be honestly laid alongside another. Last, compiling means putting out an early estimate that a later and better one will replace.

The National Statistical Office measures and holds neither lever. The office is therefore a third kind of body in the arrangement, not a junior version of the other two. The distinction is easy to say and easy to forget. The numbers the office produces are the numbers everybody argues about, and the arguing happens where the deciding happens. A number that is quoted in a debate about a rate starts to feel as though it came from the place holding the rate.

The other half of measuring is that a first number gets replaced. An early estimate goes out when only part of the information has arrived, and a revisionThe replacement of an already published estimate by a later one, as more of the underlying information comes in. The number changing is the method working, not the method failing. follows when more of it has. How to read a release, which vintageThe stamp on an estimate saying when it was issued. Two estimates covering exactly the same stretch of time can carry different stamps and different numbers. of a number is in hand and how wide the later change can be is covered separately. Read that before quoting any single figure. Only the ownership of the job matters here: the compiling and the revising both sit with the measuring body, not with anyone who might act on the result.

Why does separating the measurer from the decider matter?

Slow down here. Everything above was arrangement; this is the reason the arrangement is shaped the way it is. Imagine a single body that both pulled a lever and compiled the numbers describing what happened afterwards. Nothing about that body needs to be dishonest for the arrangement to fail. The problem is structural and it arrives before anybody's character does. A body that both decided and measured could be judged only against its own numbers, and separation is what makes a decision checkable by anyone else.

Go back to the household. Suppose the person who decides the spending also keeps the accounts book, and nobody else keeps one. At the end of the month there is exactly one record of what happened, and it was written by the person whose choices are in question. If a cousin says the month went badly, the only document available to settle it is the one the decider produced. The cousin is not calling anybody a liar. The cousin simply has nothing to hold the book against. Two books kept by two people, even two imperfect books, restore the possibility of an argument that can be settled.

Say the property precisely. Separation is easy to overstate. Separation does not make a decision correct. Separation does not make a measurement accurate. Separation does not mean the measurer and the decider never speak. All separation does is make the record of the decision and the record of the outcome come from two different places. A check requires exactly that. A check needs two things to compare. When one body supplies both, there is only ever one thing.

The property is not a compliment paid to India's arrangements, and it is not a criticism of anyone else's. Whether a given arrangement is well designed, well staffed or well funded is a different question entirely. The claim here is narrow and structural: given that the measuring is done in one place and the deciding in another, an outsider has two records instead of one. Two records instead of one is the whole of it, and enough to change how almost every macro sentence reads.

Do the two levers have to point the same way?

No, and the reason is built into the arrangement rather than sitting on top of it. Two levers act on one economy. The two levers are held by two bodies with different remits, different mandateThe job a body has been given, together with the limits that come with it. Two bodies can be doing exactly what they were each set up to do and still be acting in different directions.s and different people in the room. Nothing guarantees that the two levers agree, and that is a consequence of having two of them rather than a fault in anybody.

The household version again: two adults, two salaries, two separate bank accounts, one household. One of them is trying to build up a cushion and cuts back. The other has just been promised a bonus and commits to a bigger monthly payment on something. Neither of them has done anything wrong by their own reckoning. The household is nonetheless being pulled two ways in the same month. Two hands on one household budget rather than one hand is what creates the possibility, and no amount of good intent removes it.

Two things follow for a reader. The first is that when the two levers do point the same way, that is an observation about a particular moment and not a rule about the structure. The second is that when they point differently, that is not by itself evidence that anybody has erred. Both are permitted by the shape. Which combination is in force, and what each lever does to the economy when it moves, is covered under the levers themselves, and how the two compare as instruments is covered separately under the policy rate. Who is holding what is all that is settled here.

Where do the published records come from?

Each side publishes its own kind of document. From the measuring side come statistical releases: the compiled series, the method note that says how it was put together, and the later replacement of an earlier estimate. From the deciding side come the records of deciding: the statement of what was decided, and the reasoning the body chose to put on paper alongside it. Statistical releases and records of deciding are different kinds of document with different authors, and telling the two apart is most of what reading a macro story well consists of.

Then there is a third document, and it is the one most readers actually meet: the summary. A note, a headline, a chart with a one line caption, a thread. A summary is somebody's reading, and the record is the record. The summary is not necessarily wrong and is often useful, but it has an author, and the author selected. A good summary takes a document of many rows and presents three of them. The seven that were dropped were dropped by a person exercising judgement, and the summary does not say which seven those were.

So the habit worth building is small and cheap. When something matters, the document to consult is the one the body itself put out, from the body that performed the job in question: the measuring body for the measurement, the deciding body for the decision. How to read a release once it is in hand, including which vintage of a number it carries and how much that number can still move, is covered separately.

What can be followed through the structure without a single number?

One plain question will do. Has the general price level been rising faster lately than it was before? Walked through the three functions, it travels a long way with no figure at any step.

Step one, measuring. The National Statistical Office compiles the relevant official statistics and issues them as a release. The release tells anyone reading it what was compiled and by whom. The release is silent on whether anyone intends to do anything about it. Step two, deciding. Two bodies could act: the Reserve Bank of India on the rate side, the Ministry of Finance on the taxing and spending side. The deciding step names where the levers are. The deciding step is silent on whether either lever will be moved. Step three, publishing. The measurement lands in a statistical release and any decision lands in the deciding body's own record. Two documents, two authors, two places to go.

Not one step in that walk needed a figure, and the walk still said who does what. Learning a structure separately from learning the numbers that move through it is the point. The numbers change constantly. The three functions do not, and a reader who has the structure can go and find the current numbers at the right door instead of guessing at which body to ask.

How does an analyst or a lender actually use any of this?

In the plainest possible way: as a stop sign that fires while reading. An equity analyst reads a sector note that says a compiled statistic came from the central bank. Someone who knows the structure stops there, not because the number is necessarily wrong but because the sentence around it now cannot be trusted to have been checked, and a writer careless about the source of a figure is usually careless about more than that. A credit committee at a lender reads a borrower's submission that attributes a spending plan to the wrong body. Same stop, same conclusion about the care taken.

The habit also settles where to go when the thing itself is needed. For a measurement, the measuring body's own release is the place, and any note about it is one remove away. For a decision and the reasoning that was put on paper with it, the deciding body's own record is the place. A household reading a headline about the cost of living runs that same check at a smaller scale, simply by asking who actually produced the number sitting in it. The practical value of the structure is that it names which door to knock on before any number is in hand.

What does knowing the structure not settle?

A map of who does what is silent on every question about outcomes. The map does not say what any body will do. Nobody has taken those decisions yet. Nor does the map say whether the arrangement is a good one. Whether an arrangement is well designed is an assessment of an institution rather than a description of one. And a map of the three functions names no figure at all. A structure that had to be reprinted every time a number moved would not be a structure.

The map gives something durable instead. Any macro sentence can be met with three questions: which of the three functions it concerns, which body performs that function, and where the underlying document would be. Three questions that survive every revision, every reorganisation and every change of number are a great deal to get without naming a single figure.

The household keeping an accounts book on the shelf and the two adults with two salaries and one budget are invented.
Educational material. Not advice on any investment, tax, budget or market position.