IIT JAM Economics common mistakes
IIT JAM Economics · Understand the traps and how to avoid them.
Attempting microeconomics numericals (elasticity, consumer/producer surplus, cost curves) by plugging numbers into a memorised formula without first identifying the underlying functional form or drawing the geometry of the problem.
Candidates learn formulas as isolated strings (e.g., point elasticity = (dQ/dP)(P/Q)) and then apply them to non-linear or discrete demand data where the derivative is not constant, or confuse arc elasticity with point elasticity. Under time pressure they skip the setup step.
What to do instead
Before computing, write one line stating the functional form and the point of evaluation (e.g., 'linear demand, elasticity at P=10, use derivative -b'). Practice 20 problems where the same elasticity question is posed with linear, isoelastic, and tabular demand so the choice of formula becomes a decision, not a reflex.
Treating 'increase in demand' and 'increase in quantity demanded' as interchangeable in supply-and-demand shift questions, and shifting the wrong curve or both curves.
Everyday language blurs movement along a curve with a shift of the curve. Candidates also misread the shock: a price change moves along the curve, a non-price determinant shifts it. Under exam pressure the distinction collapses.
What to do instead
Adopt a two-column routine: 'price change -> movement along' vs 'determinant change (income, tastes, prices of related goods, technology, input costs, expectations, number of buyers/sellers) -> shift'. For each practice question, explicitly name which determinant changed and which curve it belongs to before drawing.
In macroeconomics, computing multipliers or equilibrium output without checking whether the economy is at or below potential output, or applying the simple Keynesian multiplier when prices are assumed flexible.
Textbook multiplier formulas (1/(1-MPC), 1/(1-MPC(1-t))) are memorised without their assumptions. Candidates also forget that the balanced-budget multiplier equals 1 only under specific conditions, and mix up the tax multiplier sign.
What to do instead
For every multiplier problem, write the model assumptions first (fixed prices, closed/open economy, lump-sum vs proportional tax). Then derive the multiplier from the equilibrium condition Y = C + I + G + NX rather than recalling it. Do a set of problems varying one assumption at a time.
In statistics, using the population standard deviation formula or the z-distribution when the sample is small and the population variance is unknown, and misreading one-tailed versus two-tailed critical values.
Candidates memorise 'use t when n < 30' as a rule of thumb without the real condition (unknown population variance), and they default to the familiar z-table. One-tailed vs two-tailed is decided by the wording of the alternative hypothesis, which is often skimmed.
What to do instead
Write H0 and H1 explicitly in symbols before choosing a test; the inequality direction in H1 fixes the tail. State whether sigma is known. Build a one-page decision tree (known sigma -> z; unknown sigma -> t with df = n-1) and apply it to 15 mixed problems.
Confusing correlation with causation and misinterpreting regression coefficients, R-squared, and p-values in interpretation questions.
Candidates learn the mechanics of OLS but not the language of interpretation. They say 'a coefficient of 2.5 means Y increases by 2.5' without the 'holding other variables constant' clause, and they read R-squared as a measure of causal fit or of coefficient significance.
What to do instead
Practise writing interpretations in a fixed template: 'A one-unit increase in X is associated with a b-unit change in Y, holding other regressors constant.' Separately drill what R-squared does and does not measure, and what a p-value is (probability of data at least this extreme under H0, not probability H0 is true).
In microeconomics, solving utility-maximisation or cost-minimisation problems without setting up the Lagrangian or the tangency condition, and ignoring corner solutions and the second-order condition.
Candidates jump to MRS = price ratio as a formula and forget it is a first-order condition valid for interior optima. They miss cases where the optimum is at a boundary (perfect substitutes, non-negativity constraints) and never check whether the stationary point is a maximum.
What to do instead
Always write the Lagrangian, list first-order conditions, then explicitly test for corner solutions by comparing MRS with the price ratio at the boundary. For Cobb-Douglas, verify the second-order condition or argue quasi-concavity. Do a set contrasting perfect substitutes, perfect complements, and Cobb-Douglas.
Mixing up real and nominal variables, and misapplying the GDP deflator versus the CPI in inflation and growth calculations.
Both are price indices, so candidates treat them as interchangeable. They forget the deflator covers all domestically produced goods while the CPI covers a fixed consumption basket including imports, and they confuse base-year weighting with current-year weighting.
What to do instead
Write the definitions side by side with the formulas (deflator = nominal/real x 100, Paasche-type; CPI = Laspeyres-type fixed basket). Then do problems where the two indices diverge because import prices or the basket change, and state which index the question implies before computing.
In international trade and open-economy macro, misapplying purchasing power parity, interest rate parity, and the Marshall-Lerner condition, or confusing nominal and real exchange rates.
The direction conventions of exchange rate quotes (domestic per foreign vs foreign per domestic) flip the algebra, and candidates memorise parity equations without the arbitrage logic. Marshall-Lerner is quoted as a condition on elasticities but applied to the wrong variable.
What to do instead
Fix one quotation convention and stick to it in every problem, writing it at the top of the page. Derive each parity condition from the no-arbitrage argument once, then re-derive it under the exam convention. For Marshall-Lerner, state explicitly that the sum of export and import demand elasticities must exceed one for a depreciation to improve the trade balance.
In development and growth economics, quoting Solow or Harrod-Domar results without identifying steady state versus transition dynamics, and misattributing convergence predictions.
Candidates memorise 'saving raises growth' without the distinction between level effects and growth-rate effects, and they conflate absolute and conditional convergence. The Solow steady state is a level, not a permanent growth rate.
What to do instead
For each model, write what is endogenous, what is exogenous, and what happens in steady state versus along the transition. Practise stating: in Solow, a higher saving rate raises the steady-state level of output per worker but not the long-run growth rate; unconditional convergence holds only if countries share the same parameters.
Spending disproportionate time on one long numerical question early in the paper, leaving later sections unattempted, and leaving multiple-choice items blank because negative marking is assumed.
Candidates treat the paper as a school test where every question must be finished. Without a timing plan and without confirming the actual marking scheme from the official information brochure, they either guess wildly or omit answers out of fear.
What to do instead
Before the exam, confirm the marking scheme from the official IIT JAM information brochure and build a section-wise time budget with a hard stop per section. In practice mocks, enforce the stop and log which question types overran. Attempt every item if there is no negative marking; if there is, apply a threshold rule decided in advance.