Argentina vs. Australia: how two near-identical economies came apart
In 1900 they shipped the same beef to the same buyer and earned roughly the same wage. A century later one is three times richer. The reason is not luck, and it is not character.
Compare ARG and AUS on the GDP mapThe symmetric start
“In 1913, Argentina was richer than France, Germany, and Italy. It was one of the ten wealthiest countries in the world. A hundred years on, that sentence reads like science fiction.”
— the recurring framing of “the Argentine paradox,” as it circulates in The Economist and a hundred YouTube explainers since
In 1900, Buenos Aires was one of the richest cities on Earth. Its opera house rivalled Europe’s, its income per head ran ahead of France and Germany on some measures, and British capital was laying its railways as fast as it was laying Australia’s. An Australian and an Argentine of that year would have recognized each other’s economies on sight: temperate grassland, beef and wheat and wool, European migrants stepping off ships, all of it pointed at the same market in London. Today one of them is three times richer than the other. This walkthrough is about which one, and why.
Start with what made the two economies twins, because the resemblance was real and it is the reason the divergence is a genuine puzzle rather than a foregone conclusion. Both belonged to what economic historians call the regions of recent settlement: temperate, sparsely populated, European-settled territories that grew rich in the half-century before 1914 by exporting land-intensive primary commodities into a single integrating Atlantic economy. Land was abundant and labour and capital were scarce, so both imported the missing factors (European migrants on assisted passages, British capital through London bond markets) and shipped wheat, beef, and wool back the other way. The engine of growth in such an economy is its staple export, and the load-bearing variable is the price that staple commands abroad.
Stripped to its core, a settler economy’s real income per head tracks the terms of trade for its staple. If $P_x$ is the export (staple) price, $P_m$ the price of imported manufactures, and the country exports a roughly fixed land-intensive bundle, then real purchasing power moves with
$$\tau = \frac{P_x}{P_m}$$When $\tau$ rises, as it did across the 1870–1913 wheat-and-wool boom, the settler economy gets richer for free. The vulnerability is the mirror image: a staple economy lives and dies by a price it does not set.
A settler economy gets rich by selling one kind of thing (grain, meat, wool) and buying everything else. So its prosperity rides on one number: how much the world will pay for what it sells, relative to what it has to buy. When that ratio is high, the country is rich without having to be clever. When it falls, the whole model is exposed. Argentina and Australia both ran on this machine, and both ran it well.
So far the two are interchangeable. Now the first complication, and everything that follows turns on it: the shared model concealed two structural differences that the matching income number hid. The first was institutional inheritance. Australia received a more complete transplant of British practice: common-law property rights, parliamentary norms, and an early and wide settler franchise. Argentina’s constitutional order was real but younger, its franchise narrower and more contested until the 1912 reform. The second was the political position of the export elite. Argentina’s landed oligarchy, the estancia owners of the pampas, was unusually concentrated and politically dominant, in a way Australia’s pastoralists never were, because the Australian pastoralist had to share the polity with an organized, enfranchised urban working class almost from the start. The income twins were not institutional twins.
The formal home of the staple-and-terms-of-trade growth apparatus is the growth-theory chapter, and the wider settler-economy growth trajectory (why these economies converged on the rich world and then, in Argentina’s case, fell back) is the opening of the development-economics chapter. The history book carries the chronology of the settler boom itself: the 1820–1929 globalization era in which Britain capitalized the regions of recent settlement, and the contrast it draws between settler and extractive trajectories, sit in Economic History Ch.10 §10.5.
Before going further, look at the two trajectories on one axis. The GDP-per-capita map carries both Argentina and Australia in its annotated set, with the year-events that mark each turn. Lay them side by side and the shape of the whole argument appears: near-convergence around 1900, then a slow scissoring apart that no single bad year explains.
“Argentina was never really a rich country. It had a rich elite, a European facade, and a culture that could never sustain prosperity. The decline was written in long before the coups.”
— the fatalist reading, as it recurs across popular explainers of “why Argentina failed”
“Argentina was always going to decline.”
The most popular story about Argentina is that the fall was inevitable — a matter of national character, or a wealth that was always a mirage. It is the one frame that makes the comparison pointless before it begins. So it is worth dismantling first.
How symmetric was the start, really?
“Same temperate grassland, same staples, same buyer, same migrant streams, same London capital — and within a few percent, the same income. If two economies were ever a controlled experiment, this was it.”
— the “essentially symmetric start” framing, the premise that makes the divergence worth explaining
This framing is the reason the comparison is powerful, and it is mostly right. Hold the staple-export structure, the British nexus, the climate, and the settler demographics fixed across both cases, and you have controlled for nearly everything a single-country story leans on. Whatever explains the divergence has to be something that differed despite all that sameness — which is exactly what makes Argentina-and-Australia a natural experiment rather than two unrelated histories. Take this voice seriously and you cannot fall back on “different commodities” or “different luck”: the structure was shared, so the explanation has to live elsewhere.
“Equal incomes, unequal institutions. The pampas were owned by a few hundred families who ran the state; the Australian squatter answered to a parliament and a labour movement. That difference was already there in 1900 — you just couldn’t see it in the GDP figure.”
— the “asymmetric inheritance” correction, argued at full strength
This voice is also right, and the two are not in contradiction — they describe different layers. The incomes matched; the political structures did not. Australia’s export wealth flowed through a settler democracy with an enfranchised working class and an emerging arbitration machinery; Argentina’s flowed through a landed elite that dominated the state and faced no comparable counterweight until Peronism erupted decades later. The correction matters because it tells you where to look for the divergence: not in the commodities, which were shared, but in who held power and how durably the rules constrained them. That is a difference of degree in 1900 (Argentina was a real republic, not an autocracy), yet it is the seam along which the two economies tore.
Where this leaves us
The starts were close enough to make the divergence a genuine puzzle. The income number was not a mirage, and the staple-export model was truly shared — so the “identical twins” framing is half right, and the half that is right is what gives the comparison its force. But beneath the matching income, the two economies already differed in two things that would turn out to matter more than the commodities ever did: who held political power, and how strong the institutions were that constrained them. The half of the twins framing that is wrong is where the rest of this walkthrough lives.
If the commodities were the same and the income was the same, the explanation has to live where the two countries were already different — or where they diverged after. The first place to look is the one Australia got conspicuously right and Argentina got conspicuously, repeatedly wrong: who gets to govern, and whether the rules survive the people who lose the vote.
The arbitration state vs. the coup cycle
“Argentina is the only country in the world that, having reached the level of a developed nation, then receded to underdevelopment. It moved in and out of the club of rich, inclusive societies — and the institutions never held the door open.”
— after Daron Acemoglu & James Robinson, Why Nations Fail (2012), on Argentina’s repeated round trips through inclusive institutions
Australia held elections and honored their results for a century. Argentina had its government overthrown by the military in 1930, 1943, 1955, 1962, 1966, and 1976. One settler economy built a durable machine for settling distributional fights at the ballot box and the arbitration bench; the other settled them, again and again, by force. Start with the contrast at its sharpest and ask how much of a three-fold income gap it can carry on its own.
The apparatus here is institutions as the fundamental cause of long-run growth, the framework Acemoglu, Johnson, and Robinson built and identified through the settler-mortality natural experiment. The claim is that the deep driver of why some societies compound growth and others do not is the quality of their institutions: secure property rights, real constraints on the executive, and broad political participation. Where rulers can be checked and power changes hands by rule rather than by force, investors and citizens commit for the long run; where the executive can expropriate or be overthrown at will, they do not. Both Argentina and Australia are textbook “regions of recent settlement,” so the natural-experiment logic applies directly — but the institutions that actually took root in the two places differed in exactly the dimension the theory says matters most.
The identification problem is that institutional quality $I$ and income $Y$ are jointly determined — rich countries can afford good institutions. The AJR move is to instrument $I$ with a variable that affects $Y$ only through institutions:
$$Y_i = \alpha + \beta\, \hat{I}_i + \varepsilon_i, \qquad \hat{I}_i = f(\text{settler conditions}_i)$$where settler conditions shaped whether colonizers built inclusive or extractive institutions. The estimated $\beta$ is large: institutions, not geography or culture, carry most of the long-run variance.
Strip out wealth, geography, and culture, and the thing that best predicts which countries got rich and stayed rich is whether their rulers could be reliably checked — whether power changed hands by rule rather than by force. Australia could; Argentina, for most of the twentieth century, could not. The theory says that single difference should carry a lot of weight. The hard part is what it leaves unexplained.
Apply it and the gap starts to make sense. Australia’s executive was reliably constrained and its governments changed by election for the whole century; Argentina’s constraint on the executive repeatedly collapsed, with the army as the recurrent veto player that simply removed governments it disliked. A country that cannot keep its rules for longer than one government cannot sustain the long-horizon commitments (to property, to contracts, to a coherent policy) that compounding growth requires. That is most of a three-fold gap, and the institutions apparatus carries it honestly.
But here is where the apparatus reaches its own boundary, and the honest reader has to mark it: these institutions are partly endogenous. The hard institutions-first reading takes the coup cycle as the deep cause and stops there. Yet the coup cycle had a driver of its own: a distributional conflict, between the agro-export elite and the Peronist labour movement, that the institutions could not channel. “Argentina’s institutions failed” is true, but it describes the symptom; it does not yet say why they failed at the specific task they failed at. The last stage returns to collect that debt. The formal political-economy of why institutions constrain, or fail to, is the institutional-economics chapter; the AJR framework itself is its §18.3.
The intellectual lineage of this answer, the institutionalist tradition running from Veblen through North and Williamson’s New Institutional Economics to the Acemoglu–Robinson revival, is the home of History of Economic Thought Ch.15 (The institutionalist tradition); you can trace the school as a connected cluster in the institutional view of the lineage graph. And this is the worked, matched-pair version of a question the sibling walkthrough “Why are some countries rich and others poor?” takes at the general, cross-country altitude: that walkthrough builds the AJR apparatus in the abstract; this one applies it to the one pair where the staple-export structure is held fixed.
“Nations fail today because their extractive economic institutions do not create the incentives needed for people to save, invest, and innovate. It is all, in the end, about institutions.”
— Daron Acemoglu & James Robinson, Why Nations Fail, 2012
“It’s all about institutions.”
The institutions-first reading is the mainstream’s central answer, and it is mostly correct. The question is whether “institutions” is the whole story or the visible surface of a deeper one.
Is the coup cycle the cause, or the symptom?
“A country that overthrows its government six times in fifty years cannot commit to anything for longer than one administration. That is why Australia compounded and Argentina did not. The coups are the whole story.”
— the institutions-first reading, at full strength
This voice is mostly right and should be argued without hedging. Compounding growth is a long-horizon game, and the precondition for playing it is a state that can keep a commitment past the next election. Australia could; Argentina, after 1930, could not. The arbitration system, the stable parliamentary alternation, the rule of law — together they were a machine for converting distributional conflict into negotiated outcomes that survived changes of government, and that machine is a genuine institutional achievement, not luck. Where the rules held, capital and skill committed for decades; where they did not, every actor played a short game, and the economy compounded accordingly. On the constrained-executive axis the institutions-first reading is close to decisive.
“The 1912 Sáenz Peña reform was a real democratization. The 1930 coup answered a genuine crisis. The breakdowns were not a defect of national character — they were an unsolved distributional problem the institutions were never built to hold.”
— the endogeneity caution, argued fairly
This voice insists the Argentine institutional sequence be read fairly rather than as self-evident culpability, and it is right to insist. The 1912 Sáenz Peña law extended the secret, compulsory, universal-male ballot: a real democratizing step, not the act of a doomed polity. The recurrent breakdowns that followed were the surfacing of a distributional conflict the institutions had no mechanism to settle: each time the agro-export elite and the urban labour bloc reached an impasse the rules could not break, the army broke it instead. Read this way the coup cycle is a symptom whose disease the next two stages diagnose — the policy lock-in it produced, and the distributional deadlock that produced it. The institutions are the level where the conflict was lost, not the reason the conflict existed.
Where this leaves us
Institutional stability is the single biggest piece of the answer, and the AJR apparatus carries it: a century of constrained executives and ballot-box transitions in Australia against a half-century of military vetoes in Argentina is most of a three-fold gap. But “Argentina’s institutions failed” is a description, not yet a cause. The institutions failed at a specific task — channelling a distributional conflict that was real, and that Australia’s institutions happened to be built to handle. Hold that thought. We collect the policy consequences first.
Unstable institutions don’t just cost you the rule of law. They cost you the ability to hold a coherent economic policy for longer than one government — which turns out to be exactly what an open trading economy needs most. While Australia spent the second half of the century cautiously opening its economy, Argentina locked itself into the opposite bet, and then could not stop paying for it.
Liberalization vs. ISI lock-in
“The prices of primary products tend to fall relative to the prices of manufactures over the long run. A periphery that stakes its future on exporting raw commodities is condemned to transfer the fruits of its own progress to the industrial centre.”
— Raúl Prebisch, the ECLA terms-of-trade-decline thesis (1950); the intellectual rationale for import substitution, stated at its strongest
After the 1930s shattered the open trading order both economies were built for, both turned protectionist. Australia eventually reversed course and opened; Argentina locked import substitution in, defaulted serially, and inflated chronically. Before judging that, take Prebisch seriously: import substitution was not a foolish idea. When the open order collapsed and would not reliably return, building a domestic industrial base behind tariffs was a rational first move that much of the developing world made. The question is why Australia could treat protection as a phase and Argentina could not.
1. ISI and the lock-in. Import-substitution industrialization runs on a familiar kit: tariffs to keep foreign manufactures out, an overvalued exchange rate to cheapen the imported machinery the new factories need, and directed cheap credit to favoured industries. The structural trap is in the arithmetic. ISI taxes the competitive export sector (the agro-export base that actually earns foreign exchange) to subsidize a manufacturing sector that stays uncompetitive, and in doing so it grows a political constituency with everything to lose from liberalization. That is the lock-in. Australia ran real protection too, through its tariff board for decades, but its export base (minerals, later services) stayed competitive and politically strong enough to force the opening when the time came. Argentina’s ISI hollowed out the export sector’s political weight and entrenched the protected one, so the constituency that could have unwound it never formed. The formal treatment of ISI and the political economy of development is the development-economics chapter’s §20.4.
2. Serial macro instability. The trade-policy choice fed a macro cycle that became Argentina’s signature. ISI manufacturing needs imported inputs, which the taxed export sector cannot finance, so the economy hits a balance-of-payments wall; the wall forces a devaluation; the devaluation feeds inflation; a stabilization is attempted; the stabilization collapses politically; repeat. You can read the rhythm straight off the country’s record (sovereign defaults in 1982, 1989, 2001, 2014, and 2020, with chronic high inflation between them), which is why the chronology itself tells the story better than re-deriving the open-economy macro models would. Those models live in the open-economy-macro chapter; the deep machinery of a single Argentine crisis, examined up close, is the sibling walkthrough “Why can’t Argentina stop inflating?”. What matters here is the recurring pattern; the sibling dissects one instance of it.
The historical record of this divergence, the Latin American debt crisis of the 1980s and the long unwind of import substitution, with the debt-to-exports trajectories of Mexico, Brazil, and Argentina laid out, is in Economic History Ch.16 §16.6. The intellectual contest behind the policy — the Prebisch–Singer structuralist case for ISI on one side, the neoclassical counter-revolution and Washington-Consensus case for liberalization on the other — lives in the development-economics lineage of History of Economic Thought Ch.16 §16.1; you can trace the structuralist and growth strands together on the growth lineage of the timeline.
“If the terms of trade move secularly against you, free trade is not a neutral default — it is a slow transfer of your growth to someone else. Building your own industry is the rational response.”
— the Prebisch / ECLA structuralist argument for import substitution
“Import substitution was a rational bet.”
Hindsight retrofits ISI as obvious folly. It was not. The failure was the lock-in and the macro accommodation — not the initial pivot, which much of the developing world made for serious reasons.
Two settler economies, two protectionisms
“Import substitution was the right response to a broken world economy. The mistake was not starting it — the mistake was a politics that could never afford to end it.”
— the Argentine policy trajectory, argued at its strongest
Argue Argentina’s path at full strength and it stops looking like stupidity. The structuralist diagnosis was real, the 1930s collapse was real, and the early industrialization ISI bought was real. What turned a reasonable bet into a trajectory of decline was not the bet itself but the lock-in: the protected industries became a constituency no government could face down, the overvalued exchange rate hardened into an entitlement, and the macro instability that followed became chronic because no political settlement could absorb the cost of fixing it. Read this way, Argentina did not “choose badly” in 1945. It chose reasonably and then found itself unable to choose again — which is a failure of political institutions wearing the costume of a policy mistake.
“Australia protected too — for decades, through the tariff board. The difference is that when protection had outlived its use, a government could be elected that unwound it and survive doing so. Hawke and Keating opened the economy and won.”
— the Australian path, argued without triumphalism
Australia is not the virtue here, and the contrast is sharper for admitting it. Australia ran heavy tariff protection for most of the twentieth century — its tariff board was no freer-trade saint, and the protected-then-opened arc took decades. What separated the two was not superior policy wisdom but institutional commitment capacity. The 1980s–90s Hawke–Keating liberalization was politically brutal and hard-fought, and it stuck because a stable government could commit to it across electoral cycles and absorb the political cost of the losers. That capacity is precisely what Stage 2 credited Australia’s institutions with, and it is the hinge of the whole comparison: same temptation, same protectionist phase, but only one country could end the phase on purpose.
Where this leaves us
The trade-and-macro-policy divergence is the proximate engine of the second-half-of-century gap. But notice what made the difference runnable. Australia could treat protection as a phase because a stable government could commit to reversing it; Argentina could not unwind ISI because no government could hold the political settlement long enough, and every stabilization attempt collapsed on the same question. The policy choices were real and consequential — and they were not free choices. They were the choices an institutionally unstable polity, caught in a distributional deadlock, could actually make.
Which brings us to the deadlock itself. Behind every failed Argentine stabilization is the same unanswerable question, who pays?, and behind Australia’s quieter century is an institutional machine built precisely to answer it. The last stage is the one that explains the other three.
Why Argentina couldn’t sustain stabilization
“The Argentine economy moves in a stop-go rhythm: a devaluation restores the external balance but crushes urban real wages; the political reaction restores wages but reignites the external crisis. Each correction sows the next. The puzzle is not why Argentina has crises — it is why it cannot stop having them.”
— after Carlos Díaz-Alejandro, Essays on the Economic History of the Argentine Republic (1970), the canonical statement of the stop-go cycle
Strip away “bad choices” and “weak institutions” as labels and ask for the mechanism. The economist who first made the Argentine paradox famous, a generation before the YouTube explainers, did not reach for national character or bad luck. He reached for a machine: two powerful sectors with opposed interests, and no institution able to broker a durable peace between them. That machine is the deepest layer of the whole comparison, and it explains the coups, the defaults, and the inflations all at once.
1. The Díaz-Alejandro stop-go cycle. Argentina had two politically powerful sectors with directly opposed interests. The agro-export elite wanted a competitive, devalued exchange rate, low domestic wages, and open trade — the conditions under which its grain and beef earned the most. The urban industrial-labour bloc, organized and mobilized under Peronism, wanted the opposite: an overvalued exchange rate to cheapen the imported inputs its factories ran on, protection for those factories, and high real wages. The two could not both be satisfied. Each devaluation that rescued the export balance cut urban real wages and detonated political backlash; each real-wage boost that bought urban peace drained reserves and forced the next devaluation. Stop-go. The cycle persisted not because anyone was stupid but because no institution could enforce a settlement that held across the swing.
2. O’Donnell’s bureaucratic-authoritarianism. This is the political-science apparatus for why the deadlock produced coups rather than mere instability. When the distributional conflict became unmanageable inside democratic institutions, the response was bureaucratic-authoritarian rule: a coalition of the military and technocrats imposing by force the settlement that democratic politics could not deliver — usually the export-elite settlement, achieved by suppressing organized labour. The coup cycle of Stage 2 is now cashed out: it was generated by the deadlock of rung 1. The endogeneity the institutions stage flagged turns out to be exactly this. The coups were not the disease; they were the recurring, violent symptom of a distributional conflict no institution could settle peacefully.
Now the move that the whole comparison has been building toward: Australia faced the same conflict. It too was a settler economy with a competitive land-and-capital export sector and an organized, enfranchised urban working class, carrying the same underlying tension between export competitiveness and domestic real wages. What differed was not the presence of the conflict but the capacity to channel it. Australia’s arbitration system, its wage boards, and its stable parliamentary alternation converted the distributional fight into negotiated, repeatable settlements — the so-called “wage-earners’ welfare state,” with the 1907 Harvester Judgment as its founding act — instead of a force-resolved swing. Same pressure; different machine. Argentina is not a country that failed a test Australia passed by virtue. It is a country that faced the same test without the institution that lets you pass it.
The development chapter’s political-economy treatment of why some societies can broker these conflicts and others cannot is §20.4. The historical chronology of the 1970s–80s stabilization failures sits in Economic History Ch.16 §16.6; the most recent instance of the same stabilization-failure pattern, examined up close, is the sibling walkthrough “Why can’t Argentina stop inflating?”.
“Argentina had every advantage and threw it away. It chose populism over discipline, printing over saving, the easy answer over the right one. Nobody did this to Argentina. Argentina did it to itself.”
— the moralizing “they brought it on themselves” verdict, in its common form
“Argentina chose badly and has only itself to blame.”
This is the frame everything so far has been working to replace. It is not wholly wrong, since the choices were real, but as an explanation it is empty, because it never asks why those were the choices on offer.
Character, or mechanism?
“Other settler economies faced hard choices and made the disciplined ones. Argentina, again and again, took the easy road and paid for it. At some point the pattern is the country, not its bad luck.”
— the moralizing frame, taken at its most serious
Take this frame at its most serious, because it is not nonsense. There really is a pattern, it really does recur across very different governments, and the recurrence really does demand explanation rather than excuse. A reading that waves every Argentine crisis away as external misfortune would be evasive; the country did make decisions, some of them avoidable, and the costs were borne by Argentines. The frame is right that the explanation must account for persistence — for why this happened over and over, under Peronists and anti-Peronists, civilians and generals alike. Any honest answer has to explain the pattern, not just the episodes. Where the frame fails is in what it offers as the explanation of the pattern: “national character.”
“The pattern is real — and it has a mechanism. Two sectors that cannot both win, and no institution able to make them share. That recurs under every government precisely because it is structural, not a matter of who happens to be in charge.”
— the distributional-conflict reading, the one argued for here
The mechanism explains exactly the persistence the moralizing frame correctly demands an account of. If the cause were one party’s bad character, the pattern would change when the party changed; it did not, because the cause is the structure both parties were trapped inside. The agro-export elite and the Peronist labour bloc each had real, irreconcilable claims, and no Argentine institution could broker a settlement durable enough to survive the next devaluation — so every government, of every stripe, ran the same stop-go and met the same wall. Australia’s arbitration state was the machine Argentina lacked, channelling the identical conflict into bargains that held. The distributional-conflict reading does what national character cannot: it explains the pattern, it explains why Australia escaped it, and it points to the thing that was actually missing — not virtue, but an institution.
The answer the comparison earns
Here is the answer the comparison earns. The divergence was produced by institutions and policy, not by commodities or character or luck, with luck carrying only a secondary share. But “institutions” and “policy” are not two rival causes; they are two levels of one cause. Argentina and Australia faced the same settler-economy distributional conflict, between a competitive export sector and an enfranchised urban working class. Australia inherited and built institutions that channelled that conflict into repeatable bargains; Argentina did not, and so the conflict resolved itself through the stop-go cycle, the coup cycle, and the serial macro crises, each stabilization foundering on the who pays question no institution could durably answer. The policy choices, ISI and serial macro accommodation alike, were the choices that deadlock could produce.
Where the mainstream still splits is the weight: how much is deep institutional quality versus the policy-and-political-economy dynamics. That disagreement is real, but it is mostly a parameter-magnitude split inside a shared frame — nearly everyone agrees institutions, policy, and their interaction explain the divergence; they disagree on the relative weights and on how much policy is downstream of institutions. At the edges sit two minority frame-level positions: a hard institutions-fundamentalist who reads policy as wholly endogenous to institutional quality, and a hard structuralist who treats the terms-of-trade and dependency structure as the frame everything else sits inside. The honest reading is joint: institutions set the menu, the distributional conflict chose the dish, and the policy choices locked in the result.
So Argentina is not a country that failed a test other settler economies passed by virtue. It is a country that drew a harder distributional hand and lacked the institutional machine to play it — which is a far more useful thing to understand, because it is the thing that travels. The lesson is not about Argentina’s soul. It is about what happens to any divided society that cannot build the institution that lets its factions lose an argument and stay in the game.
The four-stage arc
- The symmetric start, with an asymmetric structure underneath. Two settler economies near-equal in income in 1900, shipping the same staples to the same buyer — but already differing in institutional inheritance and in how dominant the export elite was. The income twins were not institutional twins, and that is where the divergence had room to grow.
- Institutional stability did most of the work. Australia’s constrained executive and ballot-box transitions against Argentina’s six coups: most of a three-fold gap. But “the institutions failed” is a description whose cause we had to keep looking for.
- Trade-and-macro policy was the proximate engine. Australia treated protection as a phase and could end it; Argentina locked ISI in and could not, because no government could hold the settlement long enough. The policy choices were real, consequential — and not free.
- The distributional conflict explains the other three. A genuine, two-sided fight between the agro-export elite and the Peronist labour bloc, with no institution able to broker it durably — the stop-go cycle that generated the coups, the defaults, and the inflations all at once. Australia faced the same conflict and channelled it; Argentina could not.
The joint reading is the one to keep: institutions and policy are not competing explanations but levels of a single one. Institutions set the menu of what a government can durably do; the distributional conflict drove which item got chosen; and the policy choices (import substitution, serial macro accommodation) then locked in the lower trajectory. Luck carried a secondary share, in commodity cycles and the timing of wars, but a pure-luck story cannot explain why the divergence persisted across so many cycles and regimes. The persistence is exactly what points back to institutions and policy, and it is why the mainstream’s remaining disagreement is mostly about weights, not about the shape of the answer.
And the comparison’s deepest payoff is its refusal of the morality tale. Argentina is not a settler economy that failed by some defect of national character while Australia succeeded by virtue. It is a country that faced the same distributional conflict every settler democracy faced and lacked the institutional machine (the arbitration bench, the durable parliament, the credible commitment) that lets a society make its factions lose an argument and stay in the game. That is the most useful thing the two trajectories teach, because it is the thing that travels to every divided society still trying to build it.