Fiscal multiplier

• 1. Introduction

  ›

  The fiscal multiplier sits at the heart of one of economics' most contentious debates: does government spending actually grow an economy, and if so, by how much? At its core, the multiplier is simply a ratio. It measures how much national income changes for every dollar the government puts into circulation. But that simple ratio carries enormous political weight. When it exceeds one, a dollar of public spending produces more than a dollar of economic output. When it falls below one, critics say the government is doing more harm than good.

  The idea was first proposed by Richard Kahn, a student of John Maynard Keynes, in 1930 and published in 1931. Kahn's insight was practical: an initial burst of spending does not vanish. It becomes income for workers, which becomes spending by those workers, which becomes income for others, rippling outward in a chain that can produce a total economic effect far larger than the original outlay.

  Yet the multiplier has never been a settled fact. The International Monetary Fund admitted in October 2012 that its own forecasting models had systematically underestimated the figure. Italian economists found values ranging from 1.4 to 2.0 under certain conditions. A study published in 2013 found the multiplier was effectively zero in economies with flexible exchange rates. The questions this documentary will explore are why the number is so hard to pin down, who benefits when it is misread, and what a misestimate of this size can cost an economy in real human terms.

• 2. Kahn, Keynes, and the Origins

  ›

  Richard Kahn published his calculations in 1931, but the political urgency behind them dates to 1929. Lloyd George, running as the Liberal candidate that year, had promised a major public works scheme to bring down unemployment. Keynes and his colleague Hubert Henderson wrote a pamphlet titled "Can Lloyd George Do It?: An Examination of the Liberal Pledge" in direct support of that proposal.

  Kahn's job was to give the argument mathematical backbone. His central claim was that much of whatever the Treasury spent on public works would return to government coffers, because a working population pays more income taxes and claims less unemployment insurance. The chain of spending and re-spending meant the net cost of public investment was lower than the sticker price.

  This theoretical scaffolding was designed to rebut the Treasury View, a doctrine holding that any increase in government spending necessarily crowds out an equal amount of private spending, leaving economic activity unchanged. Keynes and his allies regarded this as generally fallacious. The mechanism they identified was the output gap: a difference between actual GDP and potential GDP that represented idle workers and underused resources. Fiscal stimulus, amplified by the multiplier, could in theory close that gap without simply shuffling money from one pocket to another.

• 3. How the Chain Reaction Works

  ›

  Suppose a government spends one million dollars to build a factory. That sum becomes wages for builders and revenue for suppliers. The builders, now holding higher disposable incomes, spend a portion of their new earnings at local shops, on rent, and on services. Those shopkeepers and landlords, in turn, do the same.

  The total increase in GDP is the sum of every incremental rise in net income along that chain. If a builder receives one million dollars and pays out eight hundred thousand to sub-contractors, the builder's net income rises by two hundred thousand. Each subsequent recipient adds another layer, spending according to what economists call the marginal propensity to consume, or MPC. The cycle repeats until the spare capacity in the economy runs out.

  At each turn, the government also recaptures a portion of its original outlay. When money passes through a shop, purchase taxes such as VAT are collected. When workers earn wages, income tax and contributions like National Insurance in the United Kingdom are withheld almost immediately. So although the government spends one dollar, it likely receives back some fraction of that dollar through taxation, making its net expenditure smaller than the headline figure.

  The extent of this virtuous cycle depends critically on two variables: the marginal propensity to consume and the marginal propensity to import. A country where households save most of any new income, or where they spend it predominantly on foreign goods, will see a weaker domestic multiplier than one where spending stays within local markets.

• 4. Who Spends, and How Much It Matters

  ›

  MPC is not uniform across a population. Someone with above-average wealth may have an MPC close to zero, saving most of any additional income. A pensioner, by contrast, may spend every new dollar of income as it arrives, giving them an MPC of one.

  The source goes further: a pensioner or benefit recipient who gains confidence in their financial future may actually exceed an MPC of one, drawing down previously accumulated savings or borrowing to increase spending. That counterintuitive result means targeted transfers to the most financially constrained households can generate an outsized economic response.

  The nature of where that spending lands also shapes the outcome. Pensioners and low-income households tend to spend in local small businesses. Those local businesses, in turn, have their own high MPC, and their spending ripples through the same tier of local commerce. Students, parents of young children, and the unemployed are all identified as groups where spending is tightly constrained by income and where any increment will therefore pass rapidly back into the economy.

  The traditional policy recommendation that follows is to direct fiscal stimulus toward lower-income groups or toward large-scale construction projects, which carry both a high multiplier and a durable physical asset. A temporary increase in food stamps, for instance, carried an estimated multiplier of 1.73 in testimony given to Congress in July 2008 by Mark Zandi, chief economist for Moody's Economy.com, the highest figure of any policy option he evaluated that day.

• 5. Estimated Values in Practice

  ›

  Mark Zandi's July 2008 congressional testimony offered a range of multipliers across different fiscal options. Any form of increased government spending outperformed any form of tax cut in his estimates. A payroll tax holiday came in at 1.29; the refundable lump-sum tax rebates used in the Economic Stimulus Act of 2008 reached 1.26. Making the Bush tax cuts permanent ranked near the bottom at 0.29, just above accelerated depreciation at 0.27.

  The picture from Otto Eckstein's work was more sobering. Under the assumption that the money supply remains constant, the government spending multiplier fell to 0.6 and the tax-cut multiplier to 0.26. With interest rates held constant instead, those figures rose to 1.93 and 1.19. The monetary policy environment, in other words, can move the multiplier by a factor of three.

  Italian economists found values between 1.4 and 2.0 when dynamic effects were included, using mafia influence as an instrumental variable to isolate the causal effect of central funds given to local councils. That methodological detail reflects how difficult it is to isolate the multiplier from the tangle of factors that simultaneously shape any economy.

  The 2013 cross-country study added further nuance: output effects are larger in industrial than in developing countries; the multiplier is relatively large in economies with fixed exchange rates but zero in those with flexible exchange rates; open economies produce lower multipliers than closed ones; and high-debt countries also show a multiplier of zero.

• 6. When Institutions Get It Wrong

  ›

  In October 2012, the International Monetary Fund acknowledged in its Global Prospects and Policies document that its own forecasting models had been using fiscal multipliers of around 0.5. Their analysis of outcomes since the Great Recession found the actual range had been 0.9 to 1.7. The gap between the assumed figure and the realized one is large enough to change the direction of a policy recommendation entirely.

  The British case illustrates what that error can cost. The Office for Budget Responsibility used the IMF's assumptions when forecasting the economic consequences of the UK government's austerity policies. The Trades Union Congress estimated that the OBR's reliance on those underestimated multiplier values meant the damage from austerity may have been under-estimated by 76 billion pounds.

  The OBR itself, in its 2012 Forecast Evaluation Report, raised the question directly. Its language was cautious but the implication was plain: if fiscal multipliers were higher than assumed, then the tightening pursued since 2010 may have done more to depress GDP than the official forecasts ever projected.

  The American Recovery and Reinvestment Act of 2009 offered a contrasting data point. Its benefits were projected using fiscal multiplier estimates, and the period from 2010 to 2012 that followed was characterized by a slowing of job losses and growth in private-sector employment. The Economist, writing in 2009, noted that economists were deeply divided on whether such stimulus worked, partly because of a lack of empirical data from non-military fiscal expansion. The ARRA became one of the first large natural experiments to test the theory in peacetime.

• 7. Crowding Out and Its Limits

  ›

  The most persistent objection to the multiplier is crowding out. Deficit spending, the argument goes, pushes up interest rates, which discourages private borrowing for investment, reducing private economic activity by roughly as much as the government has added. Sports stadiums are a documented example where measured multiplier values have fallen below one, suggesting the public funds displaced private spending that would otherwise have occurred.

  But whether crowding out actually operates depends on the state of the financial system. When cash is being hoarded in the financial and credit system rather than flowing into productive investment, an additional supply of low-risk government securities may simply give savers somewhere to park funds they were already holding back. In that scenario, government borrowing does not displace private borrowing; the private borrowing was not happening in the first place.

  Ricardian Equivalence offers a related theoretical objection. It holds that rational households, observing government borrowing today, will save more in anticipation of higher taxes to repay that debt, reducing their own spending by exactly the amount the government has added. The net effect on aggregate demand is zero. Keynes and the multiplier tradition regard this behavioral response as unrealistic in practice, particularly during recessions when households are credit-constrained and not making long-horizon calculations.

  Even a balanced-budget stimulus, where additional public spending is fully matched by equivalent tax increases, may carry a multiplier greater than one when unemployment is high. The reasoning is that the reduction in anxiety driving cash hoarding, combined with a rise in business activity, can generate private consumption and investment gains that more than offset the tax drag. The conditions under which this holds are precisely the recessionary circumstances where the multiplier has historically been most needed.