The paradox of thrift

Detail

1. The statement

• Paradox of thrift: suppose all people in an economy try to save a larger share of their income, so the MPS rises.
• MPS (marginal propensity to save) is the share of each extra rupee of income that is saved. MPS = ΔS/ΔY.

• MPC (marginal propensity to consume) is the share of each extra rupee that is spent. MPC + MPS = 1.

• The result: total savings do not rise. They stay the same or fall, because output and income fall.

• Why it is called a "paradox": saving more is sensible for one family. It fails when everyone does it at the same time.
• The idea is linked to John Maynard Keynes, who made it popular. He pointed out that in a recession people save more to get through hard times. What the economy actually needs then is more spending and investment [5].
• Formal meaning: a rise in autonomous saving (saving that does not depend on income) lowers aggregate demand (AD). AD is the total planned spending on goods and services in the economy. Lower AD cuts output. Lower output then cuts total saving [5].

2. The trigger: economic shocks

• Economic shocks are sudden, unexpected events such as disasters, war, pandemics or sudden policy changes. They cause big changes in how people earn, spend and save.
• NCERT's example: people hear news of an imminent war or impending disaster. They become cautious, and MPC falls from 0.8 to 0.5.
• Why this counts as an autonomous cut in spending:
• The fall in MPC does not come from a change in income. It comes from an outside cause (fear).
• So it works like a cut in spending that has nothing to do with income. The multiplier then spreads it through the economy.

3. The model used

• Consumption function: C = C̄ + cY. Here C̄ = autonomous consumption (spending even at zero income) and c = MPC.
• Investment: Ī = fixed (autonomous). It does not depend on income.
• Aggregate demand: AD = C + I = Ā + cY, where Ā = C̄ + Ī (total autonomous spending).
• Equilibrium: output settles where supply equals demand, Y = AD.

• This gives Y* = Ā/(1 − c).

• Investment multiplier: k = 1/(1 − c) = 1/MPS. It shows how many rupees income changes for each one-rupee change in autonomous spending.

• NCERT numbers: C = 40 + 0.8Y and I = 10, so Ā = 50.
• Y₁* = 50/(1 − 0.8) = 50/0.2 = 250.
• Multiplier before the shock = 1/0.2 = 5. After the shock = 1/0.5 = 2.

4. NCERT arithmetic, step by step

• The shock (first hit):
• At the old income of 250, consumption falls by (0.8 − 0.5) × 250 = 75. So AD also falls by 75.
• Check: new AD at Y = 250 is 50 + 0.5(250) = 175. Output is still 250. So there is an excess supply of 75.

• Unsold goods pile up as stock (unplanned inventory accumulation). Firms cut output by 75.

• The downward rounds (the multiplier working in reverse):

• Round 1: output and income fall by 75. Consumption then falls by 0.5 × 75 = 37.5.
• Round 2: output falls by 37.5. Consumption falls by 18.75.
• Round 3: output falls by 18.75. The rounds continue and keep getting smaller.

• Total fall = 75 + 37.5 + 18.75 + … = 75/(1 − 0.5) = 150.

• New equilibrium: Y₂* = 250 − 150 = 100.

• Check: Y₂* = Ā/(1 − c) = 50/0.5 = 100.

	Before (c = 0.8)	After (c = 0.5)
Y*	250	100
C	40 + 0.8(250) = 240	40 + 0.5(100) = 90
S = Y − C	10	10
I	10	10
Multiplier 1/(1 − c)	5	2

• Savings are unchanged at 10. Income has fallen by 150 (a fall of 60%).
• Checking with the saving function, S = −C̄ + (1 − c)Y:
• Before: S = −40 + 0.2(250) = 10.

• After: S = −40 + 0.5(100) = 10.

• Why saving cannot rise:

• In equilibrium, planned saving = planned investment (S = I).
• Ī is fixed at 10, so saving must end up at 10 again.
• People tried to save more. The only effect was to push income down until saving was back at 10.

5. When savings actually fall (investment depends on income)

• NCERT's point: if investment depends on income (induced investment), a fall in Y also cuts I. Then savings actually fall.
• Illustrative example (these numbers are not from NCERT): C = 40 + cY and I = 5 + 0.02Y.
• Before (c = 0.8): Y = 45/(1 − 0.8 − 0.02) = 45/0.18 = 250. I = 5 + 5 = 10, so S = 10.
• After (c = 0.5): Y = 45/(1 − 0.5 − 0.02) = 45/0.48 = 93.75. I = 5 + 1.875 = 6.875.
• S = Y − C = 93.75 − (40 + 46.875) = 6.875.
• Saving falls from 10 to 6.875. The attempt to save more ends in less total saving.

6. The graph (NCERT Fig. 4.8)

• Old AD line: AD₁ = 50 + 0.8Y. New AD line: AD₂ = 50 + 0.5Y.
• Both lines start at the same vertical intercept (Ā = 50). Only the slope changes, from 0.8 to 0.5.
• So the AD line swings downward from the intercept and becomes flatter. This is a parametric shift in slope, not a parallel shift.
• Exam trap:
• A change in Ā (for example, autonomous investment falls) causes a parallel shift of AD.

• A change in MPC causes a swing, or rotation, of AD.

• Where AD meets the 45° line moves from E₁ (Y = 250) to E₂ (Y = 100).

7. What it means

• Fallacy of composition: the mistake of assuming that what is true for one part is also true for the whole.
• Saving more is prudent for one household.

• It becomes a collective loss when everyone does it, because one person's spending is another person's income.

• When it holds: in a demand-constrained economy with idle resources (unused machines and unemployed workers).

• There, output depends on demand. Less spending means less output.

• Contrast with other views:

• Classical / loanable-funds view:
  • More saving → more funds available to lend → interest rate falls → firms borrow and invest more.
  • Saving is turned into investment, so AD does not fall. No paradox.
  • Free-market economists such as Hayek and Milton Friedman argued that thrift is good even in hard times. Banks pass savings on as investment in new plants, skills and techniques [5].

• Long run:

  • Saving is the source of capital accumulation (building up machines, factories and infrastructure) and growth.
  • In the Harrod-Domar model, growth rate = saving rate ÷ capital-output ratio (the capital needed per unit of output). (See growth-theories-business-cycles.)
  • Thrift is harmful only when demand is short.

• Background (NCERT leec101): Keynes wrote the General Theory (1936) after the Great Depression (1929 onward). Output and jobs had collapsed, and the classical belief that markets clear by themselves had failed. The paradox of thrift comes from this demand-side view.

8. Applications

(a) COVID-19 in India (2020–21): precautionary saving

• Precautionary saving means saving more because the future feels uncertain.
• Quarterly spike (RBI):
• Household financial savings jumped to 21.0% of GDP in Q1 2020-21 (April–June, the national lockdown) [2].

• They fell back to 10.4% in Q2 2020-21, against 9.8% in Q2 2019-20 [2].

• RBI's explanation [2]:

• Households save more when the economy slows and income is uncertain.

• As the lockdown eased, households moved from an "essentials only" spending pattern back to discretionary spending (spending on things they want, not only things they need).

• Annual data (NSO, current prices) [3]:

• Household gross financial saving rose from ₹23.25 lakh crore (2019-20) to ₹30.67 lakh crore (2020-21).

• The household sector's share in the country's gross saving rose from 64.7% (2019-20) to 77.9% (2020-21).

• The paradox-like outcome at the national level [3]:

• Households saved much more. Yet India's total gross saving fell from ₹59.41 lakh crore (2019-20) to ₹57.87 lakh crore (2020-21), a fall of 2.6%.
• The gross saving rate (gross saving as a share of GNDI, gross national disposable income) fell from 29.1% (2019-20) to 28.7% (2020-21).
• GNDI itself fell by 1.5% in 2020-21.
• Caution: a large part of this fall came from heavy government dissaving (the government spent more than it earned in revenue). The general government's saving was −22.8% of gross saving in 2020-21. So it is not a pure test of the NCERT model.

(b) Later reversal (2022-23): net household financial savings fell

• Net financial saving = gross financial saving − financial liabilities (new loans taken by households).
• Worked out from NSO data [3]:
• 2020-21: 30.67 − 7.37 = ₹23.30 lakh crore, about 11.6% of GNDI.
• 2021-22: 26.12 − 8.99 = ₹17.13 lakh crore, about 7.2% of GNDI.

• 2022-23: 29.74 − 15.57 = ₹14.16 lakh crore, about 5.2% of GNDI.

• Main driver: household financial liabilities nearly doubled, from ₹8.99 lakh crore (2021-22) to ₹15.57 lakh crore (2022-23) [3].

• At the same time, household saving in physical assets (such as houses) rose to ₹34.83 lakh crore (2022-23) [3].
• The household share of gross saving came down to 60.9% (2022-23). The country's gross saving rate was 29.7% of GNDI (2022-23) [3].
• (The scaffold said "verify current RBI/MoSPI figures". These are the NSO First Revised Estimates for 2022-23, released 29 February 2024 [3].)

(c) Japan's "lost decades" (1990s onward)

• Japan had high household saving, weak demand and deflation (a continuing fall in the general price level) for many years.
• Firms and households saved and paid off debt instead of spending. Demand stayed weak even when interest rates were near zero.

(d) General lesson

• Shocks such as disasters, war, pandemics or a sudden policy change can make saving jump suddenly and deepen a slump.
• Policy response in the Keynesian view:
• When private spending falls, the government fills the gap with its own spending (counter-cyclical fiscal policy).
• Government spending (G) is autonomous, so it gets the full multiplier effect.

Prelims Hooks

• Paradox of thrift: if all households raise their MPS, total saving stays the same (Ī fixed) or falls (I induced). It never rises in the Keynesian model.
• NCERT numbers: MPC falls 0.8 → 0.5. Y* falls 250 → 100. C falls 240 → 90. S stays at 10.
• Formula: Y* = Ā/(1 − c). The multiplier falls from 5 to 2 when MPC falls from 0.8 to 0.5.
• Graph trap: a change in MPC makes the AD line swing, or change its slope, from the same intercept. A change in Ā gives a parallel shift.
• Logic trap: the paradox is an example of the fallacy of composition. It works only with idle resources / a demand-constrained economy. It is not valid in the classical loanable-funds model.
• Equilibrium condition behind it: planned S = planned I. With fixed Ī, saving is pinned to investment.
• Association: made popular by Keynes [5]. Keynes's General Theory appeared in 1936, after the Great Depression (1929).
• India data: household financial savings reached 21.0% of GDP in Q1 2020-21 (RBI) [2]. The household share of gross saving was 77.9% in 2020-21 (NSO) [3].
• Net financial saving = gross financial saving − financial liabilities. It fell to about 5.2% of GNDI in 2022-23, as household liabilities rose to ₹15.57 lakh crore [3].

Mains Points

• Saving versus demand is a matter of timing (GS-III):
• In a slump, precautionary saving shrinks AD through the reverse multiplier.
• In the long run, saving pays for investment (Harrod-Domar).

• Policy should support demand in bad times (fiscal stimulus, direct cash transfers) and encourage saving in good times.

• COVID-19 as an Indian case study [2][3]:

• Household financial saving spiked (21.0% of GDP in Q1 2020-21) while consumption collapsed.
• Total gross saving still fell (₹59.41 → ₹57.87 lakh crore), because incomes fell and the government saved less.

• This supports counter-cyclical spending, such as the government borrowing more to protect demand.

• Falling household net financial saving (2022-23) is a two-sided signal [3]:

• The good side: consumption and housing demand (physical saving) are recovering.
• The worry: households are borrowing more (liabilities ₹15.57 lakh crore). This means less domestic money is available to fund government borrowing and private investment.

• Link it to the saving-investment gap and to the current account deficit, since any shortfall must be filled with foreign money.

• Limits of the paradox: it assumes idle capacity and fixed investment.

• In a supply-constrained economy, or where higher saving lowers interest rates and raises investment (the classical view held by Hayek and Friedman [5]), thrift is good for growth.
• Good answers should state these conditions.