The competing schools of thought in economics

Mainstream economists, and some heterodox ones, accept parts of MMT's description but dispute what follows.

• The budget constraint has not gone away. Gregory Mankiw (2020) agrees a currency issuer can always pay its bills, but argues that money created to pay them is effectively borrowing once interest is paid on reserves, and that over time spending must be covered by taxes or by inflation. RBA Governor Philip Lowe made a similar point in 2020: "somebody always pays".
• Consolidation skips over real institutions. Central-bank economists, Drumetz and Pfister at the Banque de France and Beat Weber at Austria's central bank, argue that merging the treasury and central bank overlooks independence and the credibility that keeps inflation expectations anchored. Weber also credits MMT with drawing attention to neglected operational detail.
• Taxes are a slow brake. Many critics doubt parliaments would raise taxes or cut spending quickly enough in a boom; that worry is part of why inflation control was given to independent central banks.
• Currency hierarchy. Weber, and the sympathetic critics Vernengo and Pérez Caldentey, argue external constraints bite hard for countries whose currencies few people want to hold.
• Money and inflation. Mankiw argues the long-run link between money growth and inflation is stronger than MMT allows.

In a 2019 survey of leading US economists, none agreed that countries borrowing in their own currency "can finance as much real government spending as they want by creating money", and none agreed that such countries "should not worry about government deficits because they can always create money to finance their debt". MMT economists say they reject both statements as worded, because they treat inflation and real resources as the limit; critics reply that popular presentations of MMT often sound like them. On both sides, the public debate is often conducted at the level of slogans.

MMT economists reply on several points.

• Consolidation is a simplification, not a denial of institutions. Tymoigne and Wray (2013) call it "a theoretical simplification" and argue that the operational steps, such as the treasury's account at the central bank and bond auctions, do not change the outcome for a currency issuer, because the institutions coordinate closely in practice.
• Inflation is the constraint they put first. They argue MMT is the approach that makes real resources and inflation, rather than debt ratios, the test of a budget.
• External limits are acknowledged. Fadhel Kaboub, an MMT economist, argues that countries with foreign-currency debts and heavy dependence on imported food and energy have much less room, and need to build domestic capacity.
• Fiscal policy can be automatic. A Job Guarantee would expand in a slump and shrink in a boom without new legislation, answering part of the "slow brake" objection.

No country has adopted MMT's full policy framework, so the evidence comes from interpreting episodes, such as the 2020–21 stimulus and the inflation that followed, which each side reads differently.

Mainstream macroeconomics organises the economy around two forces. Aggregate demand is total spending: by households, businesses, governments and overseas buyers of our exports. Aggregate supply is what the economy can produce with the workers, equipment, materials and know-how it has. The highest output it can sustain without inflation picking up is potential output, and the difference between actual and potential output is the output gap.

When demand runs ahead of potential, firms can raise prices, and their margins, and still sell everything, and they compete for scarce workers. That is demand-pull inflation. When something raises the cost of production, such as an oil price spike or floods, supply falls and prices rise even if demand hasn't changed. That is cost-push inflation.

The RBA's explainers use exactly this framework. It matters for policy because the main tools act on demand. Interest rates and budgets can slow or speed up spending within months; they cannot quickly create more workers, power stations or houses. So in the mainstream view, a temporary supply shock can be "looked through" if expectations stay anchored, but demand may need slowing if inflation threatens to persist.

Potential output is not fixed. It grows with population, investment and productivity, and it can only be estimated, never observed.

MMT and post-Keynesian economists use much the same picture: MMT's "real resource constraint" is close to the mainstream's potential output. The differences are of emphasis. They argue economies usually operate with spare capacity, visible in unemployment and underemployment, so extra spending often raises output rather than prices. They are sceptical of output-gap estimates, which are often revised heavily after the fact. And they argue supply depends on demand: strong demand draws people into work and encourages investment, while long slumps destroy skills and capacity. Post-Keynesians add that cost-push forces, including market power and conflict over income shares, drive much real-world inflation.

Prices and wages are set by people looking ahead. A business that expects its costs to rise 5 per cent next year will plan to lift its prices; workers who expect 5 per cent inflation will push for pay rises to match. So expectations of inflation can help create it.

The mainstream key is whether expectations are anchored. If households and businesses assume inflation will return to around the 2.5 per cent midpoint whatever happens this year, a temporary shock fades. If they come to expect high inflation to continue, it can become self-sustaining, and bringing it down later costs more lost output and jobs. Many economists drew that lesson from the 1970s.

Expectations also matter for interest rates. What affects borrowing and spending is the real interest rate: the interest rate minus expected inflation. A 4 per cent interest rate with inflation expected at 6 per cent is a real rate of minus 2 per cent, which still encourages borrowing.

This is why the RBA publishes a target, explains its decisions, and watches surveys of consumers, businesses, unions and market economists, plus market prices that imply expected inflation. Its explainer says expectations have been consistent with the target since the early 1990s. In February 2023 it said that medium-term expectations "remain well anchored, and it is important that this remains the case".

MMT and post-Keynesian economists generally give expectations a smaller, less independent role. They see inflation as driven by concrete pressures: spending beyond capacity, supply shocks, indexation of wages, contracts and administered prices, and bargaining power over wages and margins. Expectations matter mainly through those channels, for example when contracts are linked to past inflation.

Some mainstream economists share the scepticism. Jeremy Rudd, a Federal Reserve Board economist, argued in 2021 that the theory and evidence behind the belief that expectations drive inflation are weak, and that acting on it unquestioningly could lead to serious policy mistakes. Measurement is a problem too: the RBA's own review of the measures notes that consumers' expectations do not line up well with actual inflation, and market measures include risk premiums.

In 1958 the economist A. W. Phillips found that, in nearly a century of UK data, wage growth was faster when unemployment was low. Economists soon extended this to prices: a tight labour market means faster inflation. For a while it was treated as a menu: choose lower unemployment and accept more inflation.

Friedman and Phelps argued that the menu was an illusion in the long run. Hold unemployment low and people come to expect higher inflation and build it into wages and prices, so the curve shifts up. The modern mainstream version is the expectations-augmented Phillips curve: inflation depends on expected inflation, on how tight the economy is, and on supply shocks.

From the 1990s to 2019 the curve looked flat in many countries: unemployment fell with little rise in inflation. Mainstream explanations include well-anchored expectations, measurement problems, and central banks themselves leaning against inflation, which hides the relationship in national data. Hooper, Mishkin and Sufi (2019) found a clearer, nonlinear relationship in US state and city data, suggesting the curve was hibernating rather than dead. Bernanke and Blanchard's study of the US pandemic-era inflation found tight labour markets becoming a more important driver of inflation from 2022, as the earlier price shocks faded.

Post-Keynesian and MMT economists generally accept that inflation can rise when the economy runs very hot, but doubt the Phillips curve is a stable relationship that policy can use. James Galbraith (1997) argued that the evidence for a precise trade-off is weak and that the concept had become a poor guide to policy. Many heterodox economists read the flat curve of the 2010s as evidence that economies had more room than central banks assumed. They also argue the relationship depends on institutions, such as union strength, wage indexation and import competition, rather than being a natural law, and that it says nothing about who bears the cost of the unemployment.

The RBA defines the NAIRU, the non-accelerating inflation rate of unemployment, as "the lowest unemployment rate that can be sustained without causing wages growth and inflation to rise". It comes from the expectations-augmented Phillips curve: if unemployment stays below the NAIRU, wage and price growth keep picking up; above it, they ease.

The RBA uses the gap between unemployment and its NAIRU estimate as one input to forecasting wages and inflation, alongside underemployment, job vacancies and business surveys. It is open about the uncertainty. The NAIRU cannot be observed; it is inferred from models. A 2017 RBA estimate was about 5 per cent, with a 70 per cent confidence band of roughly 4 to 6 per cent, and later research put it below 5 per cent before the pandemic. In February 2024 the RBA's models put the unemployment rate consistent with full employment at about 4 to 5½ per cent. Estimates in Australia and overseas have drifted down for decades.

Two points are often missed. First, the NAIRU is not a target: the RBA's mandate is low, stable inflation and full employment. Second, the RBA says monetary policy cannot lower the NAIRU itself. That depends on skills, training, job matching and bargaining arrangements, so "policymakers must look beyond monetary policy".

Post-Keynesian and MMT economists are among the NAIRU's sharpest critics. James Galbraith's 1997 paper "Time to Ditch the NAIRU" argued its theory and evidence were weak. Bill Mitchell and Pavlina Tcherneva argue that the NAIRU framework uses unemployed people as the economy's inflation buffer: when inflation threatens, policy deliberately slows hiring. Because estimates are so uncertain and keep being revised down, they argue, steering by them risks holding unemployment needlessly high for years. Their alternative buffer is a Job Guarantee: in a slump, people move into public jobs at a fixed wage rather than into unemployment, which they argue would anchor wages without idle workers. Critics of the Job Guarantee question its cost, administration, the value of the work, and whether it would really stabilise prices.

The RBA's explainer names four channels through which interest rates affect spending:

• Saving and investment. Higher deposit rates reward saving; higher loan rates make borrowing for homes, cars and business projects dearer, so fewer go ahead.
• Cash flow. Borrowers with variable-rate loans pay more interest and have less to spend. Savers earn more on deposits, which partly offsets this, but the RBA judges the borrower effect larger. In 2017 about four-fifths of Australian housing loans had variable rates, so this channel works faster here than in countries with long fixed-rate mortgages.
• Asset prices and wealth. Higher rates tend to lower house and share prices, so people feel less wealthy and have less collateral to borrow against.
• Exchange rate. Higher rates relative to other countries tend to lift the dollar, making imports cheaper, which directly lowers inflation, and exporters less competitive.

Alongside these, a credible target is meant to anchor expectations, so wage and price setters believe inflation will return to target.

Lags and evidence. The RBA says the maximum effect takes one to two years, and that "there is a large degree of uncertainty about the timing and size of the impact". A 2025 RBA review of its models found that a one percentage point change in the cash rate has a peak effect of about ¼ to 1 per cent on GDP and ⅛ to ½ a percentage point on inflation, after about one to two years. The exchange rate accounted for around one-third to two-thirds of the inflation response, and housing was especially sensitive. The household cash-flow channel, though large for individual borrowers, was small in aggregate, because borrowers and savers partly offset each other.

These model results are not the only evidence. Studies that isolate surprise changes in policy rates, such as Christina and David Romer's for the United States, find that tightening lowers output and, after a longer lag, inflation. The Australian evidence of this kind is less clear-cut: some statistical studies find prices briefly rising after a tightening, the so-called price puzzle, and RBA researchers have not been able to remove it in Australian data. Mainstream economists also point to episodes such as the US disinflation of the early 1980s and Australia's in the early 1990s, when tight policy was followed by a lasting fall in inflation, at a large cost in jobs.

MMT economists don't deny that higher rates can slow spending; they doubt it is a reliable or fair way to control inflation. Randall Wray gives three reasons MMT does not support using rates to manage demand: there is little evidence that spending is very sensitive to interest rates; the evidence that rate changes move inflation is weak; and when rate rises do bite, it is often through falling asset prices, insolvencies and financial stress. MMT prefers a low, stable policy rate and wants fiscal policy, including automatic stabilisers such as a Job Guarantee, to manage demand.

Two further criticisms are common. Distribution: rate rises work by squeezing mortgage holders and slowing hiring, while people with savings earn more; Bill Mitchell has argued their effects fall hardest on low-income earners. Interest income: higher rates mean more interest paid to bondholders and depositors. Warren Mosler argues this could offset the squeeze or even add to inflation; Stephanie Kelton presents it as an open question about who receives that income and how much they spend.

Post-Keynesians add that rates fit supply-driven inflation poorly: they cannot create energy or shipping capacity, and work only by weakening demand enough to offset the shock.

A fiscal multiplier measures how much GDP changes for each dollar of extra government spending or tax cuts. A multiplier of 1.5 means $1 billion of extra spending raises GDP by $1.5 billion; a multiplier of 0.5 means it raises GDP by only $500 million, because something else falls.

Mainstream economists see the multiplier as varying with circumstances. IMF reviews find it tends to be larger when:

• less of the spending leaks abroad through imports (Australia is a fairly open economy);
• the central bank does not offset it by raising rates, for example when rates are already near zero;
• the economy has spare capacity, as in a recession;
• people are confident the government's finances are sustainable.

In normal times, the mainstream expectation is that the RBA will lean against a fiscal boost that adds to inflation, which pulls the multiplier down.

The euro crisis shifted views. Olivier Blanchard and Daniel Leigh, then at the IMF, found in 2013 that forecasters early in the crisis had assumed multipliers well below the true ones, so planned budget cuts did more damage to growth than expected.

MMT and post-Keynesian economists generally expect multipliers to be larger than mainstream models assume, because they think economies usually have spare capacity and because they reject the idea that deficits push up interest rates by competing for a fixed pool of saving. In Wray's words, "Government deficits do not drive interest rates up." Spending creates income, which creates saving, so public spending need not crowd out private investment unless the central bank responds by raising rates, which they see as a policy choice. They also stress what the spending buys: public investment in skills or infrastructure can raise future capacity. And the multiplier works in reverse: cutting deficits in a weak economy shrinks incomes.

In the mainstream and official description, the Australian Government finances its deficits by borrowing. The Australian Office of Financial Management (AOFM) issues Treasury Bonds and Treasury Notes, mostly through competitive tenders in which registered bidders bid the yields; tenders made up about 80 per cent of annual issuance when the AOFM described them in 2019. The AOFM's stated role is to ensure the government can meet its spending, investment and debt payment obligations, and it forecasts daily cash flows so the government's accounts at the RBA have funds.

Yields are set in the market, influenced by the expected path of the cash rate, global bond yields, inflation risk and investor demand. Because the government borrows in its own currency, its bonds are treated as free of default risk, and the government yield curve is the benchmark for other borrowing.

Mainstream economists see value in this arrangement. Markets put a price on public borrowing, and the RBA does not finance the budget directly (the government's accounts there have only a strictly limited overdraft), which keeps monetary policy separate from budget decisions. When the RBA bought government bonds in 2020–22 it did so for monetary policy reasons; its own review estimated the program lowered long-term yields by about 30 basis points, at an estimated cost to the RBA of $35 to $58 billion to 2033.

In 2020 Governor Lowe said monetary financing was "not an option under consideration in Australia, nor does it need to be", because the government could borrow cheaply in the market.

MMT economists describe the same flows differently. Viewing the treasury and central bank together, they argue, government spending adds reserves to the banking system and bond sales then swap those reserves for an interest-paying asset. On this view bonds don't fund spending: they give savers a safe asset and, historically, helped the central bank hold its interest rate target. Wray calls bond sales "not really a borrowing operation". Because the RBA now pays interest on settlement balances, MMT economists argue the government could issue fewer bonds, and that the rate it pays is ultimately a policy choice. Bill Mitchell goes further and argues bond issuance is unnecessary for a currency issuer. None of this disputes that, under current Australian law and practice, the AOFM issues bonds and the government spends from accounts that must hold funds; the argument is about what that arrangement is for.

Mainstream economists judge fiscal sustainability mainly by the debt-to-GDP ratio: government debt as a share of the economy's annual income. The Parliamentary Budget Office treats the budget as sustainable if that ratio is expected to be stable or trending down. Its September 2025 outlook projected gross debt falling from 35.5 per cent of GDP in 2025–26 to 31.8 per cent in 2035–36, with debt-to-GDP trending down in 23 of 27 long-term scenarios. (The May 2026 Budget, using later GDP figures, put gross debt at 33.1 per cent of GDP in 2025–26 and 34.0 per cent in 2026–27.) The government's own strategy is to reduce gross debt as a share of the economy over time.

The key arithmetic is r versus g: the interest rate on debt compared with the economy's growth rate. If growth exceeds the interest rate, debt can shrink as a share of GDP even with modest deficits. If the interest rate exceeds growth, the government needs primary surpluses (revenue above spending, before interest) to stop the ratio rising.

Why worry, if Australia borrows in its own currency? Not about default. Mainstream economists worry that interest costs crowd out other spending or require higher taxes later; that interest rates can rise, as they did from 2022, lifting interest costs; that high debt leaves less room for the next crisis; and that doubts about a government's commitment to low inflation raise borrowing costs and inflation risk. Olivier Blanchard (2019) argued that while r stays below g, debt may cost less than often assumed, but warned conditions can change.

MMT economists see the debt ratio as the wrong headline. For a government borrowing in its own currency, they argue, the debt is the record of past deficits, held as safe savings by households, super funds, banks and overseas investors, and interest on it is income to those holders. The meaningful test of a budget is whether total spending is pushing the economy beyond its real capacity, which shows up as inflation, not whether a ratio is rising. Because the central bank sets the short-term rate and can influence longer ones, they argue r is largely a policy choice for a currency issuer, which is one reason MMT favours keeping it low. They add that interest payments can themselves be a concern, because they flow mainly to people who already hold financial wealth.

Australia is a small open economy, and the dollar has floated since December 1983. Mainstream economists, and the RBA, see the float as a shock absorber: when commodity prices fall, the dollar tends to fall, cushioning exporters. They also see real external limits, even for a currency issuer.

• The exchange rate. Interest rates well below other countries', or a loss of confidence in policy, can push the dollar down and raise import prices. The RBA finds exchange-rate changes reach import prices quickly, and consumer prices more slowly and partly.
• Foreign funding. Australia has run current account deficits for most of the period since the mid-1970s, matched by foreign investors buying Australian assets. Their willingness to keep doing so depends partly on credible policy.
• Not all currency issuers are alike. Countries that borrow in foreign currencies ("original sin"), peg their currency, or share one, like the euro, have far less room. In a widely shared mainstream reading, the Eurozone crisis was largely a sudden stop of cross-border lending to countries without their own central bank as a backstop.

Australia's own position is unusually strong. Around three-quarters of its foreign liabilities are in Australian dollars, and it has a net foreign currency asset position, so a falling dollar actually reduces its net foreign liabilities.

MMT economists accept that a currency issuer cannot create foreign currency, and that imports must be paid for with exports, foreign investment or borrowing. They argue that for a floating issuer with debts in its own currency, like Australia, the external sector is mostly a source of shocks the float absorbs, not a binding financial constraint; foreigners holding Australian-dollar assets have chosen to save in them. They advise countries to avoid pegs and foreign-currency debt. Sympathetic heterodox critics, Matías Vernengo and Esteban Pérez Caldentey, argue that MMT underrates the balance-of-payments constraint for developing countries. Fadhel Kaboub, himself an MMT economist, argues that dependence on imported food, energy and technology is what limits much of the Global South.

In the long run, mainstream economists see living standards as set by supply: how many people work, how much capital they work with, and how productively. The RBA calls productivity growth "the main driver of higher living standards in the long run", because it lets firms raise wages without raising the labour cost of each unit they produce. Growth in this view comes from investment, education, technology, competition and good institutions.

Demand matters for the cycle: how close the economy runs to its potential. But in standard models demand does not change potential much over the long run. So the growth agenda is mainly microeconomic reform: competition, tax design, planning and housing supply, skills and infrastructure. The Productivity Commission's 2023 five-year inquiry, Advancing Prosperity, is the main recent Australian statement of this agenda.

The mainstream has softened on that separation since 2008. Janet Yellen, then chair of the US Federal Reserve, asked in 2016 whether weak demand had done lasting damage to supply, an effect called hysteresis, and whether running a "high-pressure economy" could partly reverse it. She treated this as an open research question.

Post-Keynesian and MMT economists argue that growth is demand-led for much longer than the mainstream allows. Firms invest when they expect sales; strong demand pulls people into work, encourages investment and spreads new technology; a long slump destroys skills and capacity. Nicholas Kaldor argued productivity grows faster when output grows faster. On this view, keeping the economy near full employment is itself a growth policy, and public investment can raise capacity. MMT adds that the real limits to growth are physical: workers, materials, energy and the environment, not money. Some on this side of the debate, such as Ann Pettifor, question whether endless growth should be the goal at all.

In mainstream analysis unemployment has several sources. The RBA distinguishes cyclical unemployment, from weak demand in a downturn; structural unemployment, from mismatches of skills, location or industry; and frictional unemployment, as people move between jobs. Monetary and fiscal policy can reduce cyclical unemployment; structural unemployment needs training, job matching and other policies.

The ABS counts someone as employed if they worked at least one hour for pay in the survey week, or had a job but were away from it, for example on leave. It counts them as unemployed if they had no work, actively looked in the past four weeks and were available to start, or were waiting to start a job within four weeks. Underemployment counts workers who want, and are available for, more hours. In August 2026 the ABS put seasonally adjusted unemployment at 4.6 per cent and underemployment at 6.2 per cent.

For the RBA, full employment does not mean zero unemployment: it means the most employment the economy can sustain while inflation stays low and stable, judged from many indicators. The government's 2023 Employment White Paper set the objective of "sustained and inclusive full employment", without a single numerical definition.

MMT economists argue the headline unemployment rate understates how much labour goes unused: adding underemployment gives a much larger pool of people who want more work, and some people have stopped looking. Bill Mitchell argues that in the end the federal government chooses the unemployment rate, because it can always spend enough to create jobs; mainstream economists reply that pushing too far raises inflation and that structural unemployment needs other remedies. MMT's answer to the inflation risk is the Job Guarantee: a permanent, voluntary public job offer at a living wage, which expands in downturns and shrinks as private hiring recovers. Pavlina Tcherneva argues it would end involuntary unemployment and set a wage floor for the whole economy.

Mainstream economists, and the RBA's explainers, recognise several causes of inflation:

• Demand-pull: spending running ahead of what the economy can supply.
• Cost-push: supply shocks such as energy prices, floods or shipping disruptions.
• Expectations: beliefs about future inflation feeding into wages and prices.
• The exchange rate: a falling dollar making imports dearer.

From the monetarist tradition, they also hold that high inflation cannot persist for long unless monetary policy accommodates it.

On 2021–23, mainstream research mostly found a mix. An RBA study using three methods found supply factors accounted for at least half of Australian inflation over the year to March 2023, with demand around a third in one method; the authors stress that each method has limitations. Bernanke and Blanchard found most of the 2021–22 US rise came from price shocks, with tight labour markets becoming more important later. On profits, the RBA found the rise in Australia's aggregate profit share came mostly from mining, with little broad rise in non-mining margins; Treasury's April 2023 advice to the government took a similar view. The RBA's case for raising rates was that demand was adding to supply pressures and inflation risked becoming entrenched.

Post-Keynesian and MMT economists accept demand-pull inflation, but argue many real episodes are driven by supply shocks, market power and conflict over who bears higher costs. Isabella Weber and Evan Wasner argue that in 2021–23 widely publicised cost shocks let firms with pricing power raise prices together, protecting or widening margins ("sellers' inflation"). Jim Stanford's analysis for The Australia Institute claimed excess profits accounted for 69 per cent of inflation above the RBA's target up to the September quarter of 2022. MMT economists add that indexation and administered prices spread shocks through the economy. Weber and others argue targeted tools, such as buffer stocks, price regulation and windfall taxes, can work better than slowing the whole economy.