Previously I argued that the Reserve Bank's credibility is a critical defence against a repeat of the 1970s stagflation experience. That is true, but it is only half the story. The other half is the architecture around the central bank, which determines whether that credibility can do its work.
This post is about the broader architecture necessary to manage inflation, told through the question that the set of charts below forced me to confront: Why did Australia take so much longer than everyone else to get inflation back under control after the 1973 oil shock? The UK, Japan and Italy all had an inflation peak higher than Australia's in the wake of the oil shock. All were back to pre-shock inflation levels in three to eight years. Australia took 15+ years.
A Recap on the Stagflation Mechanism
Stagflation is not just inflation and unemployment happening at the same time. Every supply shock produces some temporary version of that. Stagflation is what happens when the combination becomes self-sustaining, when the economy generates its own continuing inflation long after the original shock has passed.
The mechanism is a feedback loop. A shock raises prices. Workers, seeing their real wages fall, push for wage rises to compensate. Firms, facing higher wage costs, raise prices to protect margins. Workers see higher prices and push for another wage rise. Each round is individually reasonable. Each party is defending a position that the previous round's inflation has threatened. Collectively they generate the next round of inflation, which triggers the next round of defensive responses, and the loop runs.
What keeps the loop running is expectations. If everyone believes next year's inflation will match this year's, they price it into wage claims, supplier contracts, and rental agreements now. The belief becomes the mechanism that delivers the outcome. The original shock is long gone. The expectations it triggered have become the thing generating the inflation.
Breaking the loop requires someone to accept a real loss. Workers accepting real wage cuts. Firms accepting margin compression. Or the economy accepting unemployment as the instrument that forces both. Until someone absorbs the loss, the loop keeps running. Disinflation is not a technical problem of monetary policy. It is a problem of who bears the cost of stopping the music, and the answer to that question is always political.
That is the lens for the rest of this post. The countries that got out of the 70s quickly found a way to allocate the loss. The countries that stayed stuck, Australia most of all, had institutional arrangements that blocked every available channel for loss allocation. The plateau is what stuck looks like on a chart.
The Puzzle in the Charts
Look at the inflation peaks across the OECD after the first oil shock. Japan hit 23% in 1974. Italy peaked above 20% and did it again in 1980. The UK went above 25% in 1975 and above 20% in 1980. Australia's peak (chart at the start of this blog post) was lower at 17.5% in 1975. By the conventional measure of how bad stagflation was, Australia was not the worst case.
But look at what happened next. Japanese inflation was back below 5% by 1978. Italian inflation fell through the late 70s and reached single digits by 1984. UK inflation was below 5% by 1983. The US got there by 1982 after Volcker. Canada followed the USA. Germany never really lost control at all.
Australian inflation came down from its 1975 peak to 8% by 1978. Then it stopped. For thirteen years, from 1977 to 1990, Australian inflation sat in a band between 7% and 11%. It would not accelerate into a 1970s-style crisis, and it would not decline to anything resembling what it was before the 1973 oil shock. It just plateaued. When the break finally came, it came through the recession of 1990-91, which produced 11% unemployment and finally broke the cycle.
New Zealand's chart tells a similar story with different peaks. Inflation in the 10 to 18% range from 1974 to 1987, finally resolved by the severe recession and Reserve Bank Act of 1989-91.
So the puzzle is not why Australia's inflation was high. The puzzle is why our disinflation stalled for thirteen years when countries with worse peaks had broken the back of their inflation by the mid-80s. What was different about the Australian adjustment?
Who Carried How Much
Every OECD economy entered the 70s with some version of the post-war settlement baked into its institutions. Centralised wage determination or sectoral bargaining. Tariffs and industry protection. Regulated finance. Fixed or managed exchange rates. These were not accidents. They were the deliberate architecture of the full employment, rising real wages, expanding welfare state consensus that had taken hold across the developed world since 1945.
The question in the 1970s was how much of that architecture each country carried, and how much it was willing to throw off when the shock hit.
Germany carried the least. Sectoral bargaining with strong employer discipline, an independent Bundesbank with a clear price stability mandate, open trade integration with Europe, and a hard-currency tradition that went back to the post-war reconstruction. The Bundesbank could tighten because the rest of the machinery would not fight back. German inflation peaked at 7% in 1974 and was back to 3% by 1978. No drama. No plateau.
Japan carried something different. The machinery existed but it was adaptive rather than rigid. Shuntō, the annual spring wage offensive, was a coordinated negotiation between big employers and enterprise unions. After the 1974 shock, the major unions themselves concluded that pushing for inflation-matching wage rises would destroy the export sector and their own members' jobs. They delivered voluntary wage restraint, and the machinery that might have generated a wage-price spiral instead disarmed itself. This is culturally specific. It depends on lifetime employment, enterprise unions, and a coordination culture that would be almost impossible to replicate elsewhere. But it worked. Japanese inflation went from 23% in 1974 to below 5% in three years.
The United States carried moderate machinery and eventually produced the canonical disinflation. Wage bargaining was decentralised and less automatic than in the centralised European systems. The Fed had formal independence but had been politically pliable under Burns, who accommodated inflation through the 70s rather than resisting it. Volcker's arrival in 1979 changed the institution. Rates went above 19%. Unemployment went to 10.8% by 1982. Inflation came down from 14% to 3% in three years. The social cost was enormous, concentrated in the industrial Midwest where it produced lasting political scars, but the disinflation stuck. Volcker became the global archetype because the US held the reserve currency and because the technique worked. Every other independent central bank in the world learned from it. Canada tracked the US closely through the whole episode, with a Bank of Canada that followed the Fed's lead and a labour market that lacked the centralised indexation of the European systems. The Canadian disinflation was essentially a slightly lagged version of the American one.
The UK carried heavy machinery and a chaotic industrial relations culture. Inflation ran out of control through the 70s because no mechanism existed to restrain wage demands, and successive governments tried incomes policies that broke down. Thatcher's answer from 1979 was to accept the unemployment cost required to break the loop. UK unemployment went from 5% to 12% between 1979 and 1984. The coal strike of 1984-85 was the set-piece confrontation where the government demonstrated it would accept mass unemployment rather than concede. Inflation came down. The social cost was enormous. But it came down.
Italy carried the scala mobile, which was the most aggressive formal indexation system in the OECD. Every wage was mechanically adjusted for CPI four times a year. This guaranteed that any inflation shock would become persistent. By the early 80s it was clear the system had to go, and after a bruising political fight the Craxi government partially suspended it in 1984 and it was effectively dismantled by the early 90s. Italian inflation came down, slowly but definitively, through a combination of institutional reform and monetary tightening. Italy took its hit.
France carried heavy machinery and tried a different kind of failure first. The Mitterrand government elected in 1981 ran large fiscal expansion into a stagflationary economy, betting that domestic reflation could escape the shock. It could not. By 1983 the franc was under sustained pressure, the current account had collapsed, and Mitterrand faced a choice between leaving the European Monetary System or reversing course. He reversed course. The tournant de la rigueur of 1983 shifted France to a hard-currency strategy tied to the Deutsche mark, with disinflation delivered through the exchange rate constraint rather than through domestic political agreement. It worked. French inflation came down through the 80s, though unemployment stayed persistently higher than its pre-shock baseline. The mechanism was external discipline substituting for the domestic political will that had been absent.
New Zealand carried machinery similar to Australia's and then, from 1984, tried to throw all of it off at once. Rogernomics dismantled the tariff walls, floated the dollar, deregulated finance, abolished agricultural subsidies, and eventually in 1989 made the Reserve Bank independent with the first formal inflation target in the world. The reforms were radical and fast. The disinflation was neither. Inflation actually reaccelerated to a second peak of 19% in 1986-87 while the reforms were being implemented, and only came down decisively after the 1990-91 recession drove unemployment to 11%. The Lange government fell apart over the reforms and the New Zealand Labour Party was badly damaged for a decade. Same destination as Australia, same timing, different path, higher social cost.
And then Australia. We did not carry more lead than anyone else on any single dimension. Italy's scala mobile was more mechanical than our Arbitration Commission. The UK's industrial relations were more chaotic. New Zealand's machinery was nearly identical to ours. What made the Australian case distinctive was the specific combination, with every loss-allocation channel blocked at once.
The Australian Saddle
The Australian machinery in 1975 was unusually comprehensive even by OECD standards of the time. Each piece deserves a sentence.
The Conciliation and Arbitration Commission set wages across the economy through formal case-based determinations. By 1975 it was running explicit wage indexation, and even when indexation was formally abolished, CPI remained the dominant input to national wage cases. A supply shock that raised headline inflation was converted to base wages within a year through the arbitration process, and the higher wages then fed through to prices. The loop was tight and the institution was a century old. The Harvester judgment was 1907. Touching the system meant touching something foundational to Australian political identity.
Tariffs and quotas meant the domestic producers faced limited competitive discipline. When input costs rose, firms could pass them through to prices without losing market share to imports. The protection was extensive and it had been the centrepiece of Australian industrial policy since federation.
The financial system was regulated to the point where there was no functioning interest rate instrument. Deposit rates were capped. Lending was directed. The Reserve Bank was an arm of Treasury rather than an independent institution, and its tools were quantity restrictions rather than a flexible price instrument. Monetary policy existed, but it was blunt and constrained by its institutional subordination to Treasury. Volcker-style action was not available in the way it was available to an independent Federal Reserve.
The exchange rate was fixed, then managed against a basket. Terms-of-trade shocks could not be absorbed by the currency. They flowed directly into domestic inflation.
Fiscal policy was expansionary through most of the 70s. The Whitlam government ran large deficits through 1974-75. The Fraser government promised restraint but delivered little of it. Both governments, for different reasons, ran policies that added to the inflationary pressure rather than resisting it.
Five major channels through which a supply shock becomes entrenched inflation. All five of them open. That is the lead in the saddle that Australia carried into the 70s, and it was heavier than what almost any peer economy was carrying.
Fraser Preserved the Machinery
Fraser took office in November 1975 promising to fight inflation. He left in March 1983 with inflation at 11% and unemployment at 10%. By his own stated objectives this was comprehensive failure.
The charitable reading is that he did not have the tools. Monetary policy did not work through an interest rate the way it does now. The Arbitration Commission was not under his control. The institutional framework for what we now call inflation-fighting did not exist in Australia in 1976, and building it from scratch would have taken a political effort Fraser did not attempt.
The less charitable reading is that he declined to make the effort. The metal trades case in 1981 produced a wage explosion of around 14% in a single year. This was the moment when anyone serious about inflation would have confronted the wage system. Fraser complained about it and moved on. The tariff walls stayed up. The financial system stayed regulated. The Reserve Bank stayed subordinate. By 1983 almost none of the machinery had been touched.
The defensible reading is that dismantling the machinery quickly would have required breaking the post-war settlement, and breaking the post-war settlement was something Fraser's Liberal Party coalition was not ready to do. Business wanted industrial peace as well as low inflation. Protected industries and their workers wanted continued protection. The broader society expected full employment and rising real wages as the natural order. Fraser could complain about inflation while preserving the arrangements that generated it, and for a while that was politically comfortable.
The result was seven years of plateau. Not disaster, because the machinery never generated a runaway spiral. Not progress, because the machinery never allowed the disinflation that was happening in peer economies to happen here. Just stasis while the rest of the world moved.
Hawke Chose a Slower Path
Hawke inherited the problem in March 1983 and did something Fraser had not. He attempted to dismantle the machinery. The difference is that he attempted to do it through negotiation rather than confrontation, and that choice shaped everything.
The Accord was the central innovation. Signed with the ACTU before the 1983 election and renewed in successive iterations through the 80s, it was a corporatist bargain in which unions accepted real wage restraint in exchange for social wage expansion. Medicare was part of the bargain. So were tax cuts, family payments, and eventually compulsory superannuation. The arithmetic was that workers would accept lower real wage growth if the government delivered other things they valued.
This is a genuinely unusual policy instrument. Most countries that disinflated successfully did it by breaking labour or by having labour break itself. Australia did it by making labour a party to the adjustment. The ACTU under Bill Kelty actively participated in delivering the wage restraint the RBA would otherwise have had to impose through higher unemployment. In exchange, the Labor government built the foundations of the modern Australian welfare state. Medicare, super, the family payment system. A genuinely distinctive achievement, and one that looks better in retrospect than it did at the time.
The other reforms ran in parallel. The dollar was floated in December 1983. Financial deregulation followed in 1984-85. Tariff reductions accelerated through the late 80s and into the 90s. The Reserve Bank moved toward independence gradually, with Bernie Fraser's informal inflation targeting from 1993 and formal statement on the conduct of monetary policy from 1996. Each reform was significant. Each was fought over. Collectively they dismantled the machinery piece by piece over fifteen years.
The cost of this approach was speed. Because the Accord was delivering wage restraint slowly and because the other reforms were taking time to bite, inflation stayed in the 7 to 10% range through most of the 80s. Peer economies with more confrontational strategies were getting to 3 to 4% over the same period. Australia was not doing badly in an absolute sense. We were just doing everything slower than countries that had accepted the unemployment cost of faster adjustment.
And at the end, when the remaining disinflation had to happen, it happened through a recession anyway. Keating's tightening in 1988-89 was intended to cool an asset price boom and widen the policy gap against inflation. It overshot. The 1990-91 recession pushed unemployment to 11% and finally broke the second-round effects that the Accord had been gradually dampening. Inflation was below 3% by 1993. The gradualist strategy needed a recession to complete itself, but the recession came a decade later than it would have under a Thatcher-style approach, and the intervening decade had seen most of the institutional reform completed without the social dislocation that the UK or New Zealand had accepted.
New Zealand: Same Destination, Different Journey
Australia and New Zealand arrived at sustained low inflation at roughly the same time, around 1992-93, through very different paths. New Zealand tried to remove the lead all at once and spent seven years on a second inflation peak before the disinflation actually completed. Australia removed the lead gradually over fifteen years and held at a plateau for most of it. The outcomes on inflation were similar. The outcomes on other things were not.
New Zealand Labour was badly damaged for a decade. Australia kept Labor in government for thirteen years and used the time to build Medicare, superannuation, and a modernised economic framework that both sides of politics eventually accepted. The plateau was the price of doing the reform in a way that preserved the political coalition needed to complete it.
The harder question is whether the gradualist path was worth the price. Measured purely on inflation performance over the 1977-1990 period, Australia did badly. Thirteen years at 7 to 10% is a long time to carry elevated inflation, and the cumulative real income loss to workers over that period was significant. Measured on preservation of a functioning welfare state, avoidance of Thatcherite social dislocation, and maintenance of a political system capable of continued reform, Australia did well. Reasonable people can disagree about which measure matters more.
My own view is that Hawke and Keating got the trade roughly right for the Australia they were governing. A more confrontational strategy would have produced faster disinflation at a higher social cost, and the social cost would have included the political capacity to complete the institutional reform that eventually delivered the low-inflation equilibrium we now live in. The Accord was slow. It was also durable. The New Zealand path was fast and catastrophic for the reform coalition. The Australian path was slow and it left the coalition intact. Over fifteen years, the slower path got more done.
But the plateau was real, and it is why Australia appears at the bottom of the class if the ranking is done by speed of disinflation alone. We started the period with one of the heaviest loads and chose to remove it carefully. Careful removal takes time. The horse kept running while the lead came off, but the horse ran slowly for a long time.
The institutional story is the main driver, but it is not the only one. Australia's terms of trade cycle through the 80s worked against disinflation when commodity prices fell and the dollar weakened. The transition from fixed to managed to free-floating exchange rate in 1983 had its own complex effects on imported inflation. These forces complicated the disinflation without determining it. The institutional architecture remains the main explanation for why Australia's path looked so different from Germany's or Japan's, but the specific timing and shape of the plateau was also influenced by external cycles and regime transitions that were partly exogenous to the domestic political economy.
The Machinery Is Weakened
The lead has largely been removed. Enterprise bargaining replaced centralised arbitration through the early 90s. Tariffs came down to negligible levels. The dollar floats. The financial system is deregulated. The Reserve Bank is independent and inflation-targeting. The institutional apparatus that generated the 70s plateau is not in place any more, and the 2022-24 experience showed the weakening is real.
That episode was a genuine test. Headline CPI peaked at 7.8% in December 2022. Trimmed mean peaked at 6.8%. These are not trivial numbers. The old machinery, if intact, would have converted them into wage catch-up and persistent inflation. What actually happened was that the WPI peaked at around 4.2%. Enterprise agreement outcomes ran below CPI for most of the episode. The Fair Work Commission declined to fully index the minimum wage to headline inflation and said so publicly. Workers absorbed a meaningful real wage cut. The second-round loop did not engage.
That is encouraging evidence. The architecture built over fifteen years of reform did what it was supposed to do. The RBA tightened aggressively once it began, the wage channel dampened the pass-through, and the disinflation is on track. It is a much better outcome than 1974-75 would have produced under the old machinery facing a similar shock.
The architecture did not do all the work on its own. Post-pandemic immigration eased labour market pressure that would otherwise have pushed wages harder. Global goods disinflation through 2023-24 pulled headline inflation down from the supply side. The high share of variable-rate mortgages meant RBA tightening hit household cash flow fast, which amplified the effect of each rate rise. These were tailwinds the architecture got to work with. A future episode might come without them, and the test would be harder.
But the lead is sitting in the stable. The constituencies that liked the old system are still there. The political reflex to reach for indexation, wage protection, and fiscal cushioning under pressure has not been abolished. In 2022-23 those reflexes were visible. Pensioners wanted indexation maintained and got it. Unions wanted wage catch-up and the ACTU leadership was more restrained than their 1975 predecessors would have been, but restraint is a choice, and a different leadership facing a different shock could choose differently. The Fair Work Commission's approach is durable within the current political settlement but can be changed by legislation.
A severe enough shock would test the architecture harder than 2022-24 did. A 17% inflation print rather than 7.8% would produce political pressure of a different order. The RBA would tighten, but in a household debt environment where total household debt to GDP is around 185%, tight enough to satisfy a strict Taylor rule would run hard into financial-stability constraints. The central bank would therefore tighten less than the rule requires, inflation would stay elevated longer, and every constituency with a claim on real purchasing power would activate simultaneously. Pensioners wanting indexation. Unions wanting wage catch-up. The government wanting to cushion mortgage holders. Each claim individually reasonable. Collectively, the recipe for reactivating the machinery.
The 2022-24 episode showed the architecture can handle a moderate shock. It did not test whether the architecture can handle an extreme one, and there is good reason to think an extreme shock would be harder than the 70s were. The household leverage constraint is new. The global interconnection is deeper. The political fiscal space may be more limited. An Australia in 2026 facing a 70s-magnitude shock would not have the institutional headroom that Volcker had in 1979.
The Lead in the Stable
Australia started the 1970s with lead in the saddle, carried a particularly binding combination of it, and chose to remove it so carefully that nobody quite noticed it was happening. The horse ran slowly for a long time, but by the time the removal was done, a different economy had emerged. The slow path preserved the political coalition that could complete the reform. The fast alternatives either failed to produce faster disinflation anyway or did so at a social cost that destroyed the governments attempting them.
The reason we are not in 1975 now is that the lead has mostly been removed. The reason the current low-inflation equilibrium is more fragile than it looks is that the political economy that built the saddle has not been abolished. Under sufficient pressure, someone would pick up the lead and start putting it back on. The 2022-24 episode showed the pressure can be resisted. It did not show that the pressure can be resisted indefinitely or under any circumstance.
The first post argued that the Reserve Bank's credibility is the defence against a repeat of the 70s. This post is the complement. The credibility only works because the architecture around it has been rebuilt from what it was in 1975. The architecture only survived the rebuilding because it was done gradually enough to preserve the political coalition that completed the reform. The whole system is the product of fifteen years of deliberate choices, most of which are still in place, none of which are permanent.
The horse runs freely now because we chose to remove the lead. The choice could be reversed. The lead is sitting right there in the stable.
No comments:
Post a Comment