In the autumn of 2008 a financial crash that had begun in one corner of the American mortgage market became a worldwide contraction of credit, trade and employment. Within months, commentators across the political spectrum were announcing that Karl Marx had been proved right. Bookshops reported surging sales of his work, newspapers that rarely discussed political economy ran features on the renewed relevance of a thinker dead for well over a century, and a German finance minister was quoted observing that certain passages of the old critique read like dispatches from the present. The applause concealed a difficulty that this article takes as its subject. The Marxist tradition did not produce one explanation of the crash. It produced at least four, and the four do not agree with each other about what caused the collapse, what the long expansion of finance had been, or what sort of recovery could follow. A single event cannot vindicate four incompatible accounts at once, and the interesting work lies in setting the rivals side by side and grading each against the documented record.

That grading is the article’s method from the first page to the last. The crash itself is treated as history: a sequence of lending practices, financial instruments, institutional failures and policy responses that historians and economists have documented in detail and that can be narrated without leaning on any theory of crisis at all. Against that narrative, the rival explanations are then introduced one at a time, each in its strongest form, each asked the same questions. What does the account identify as the primary cause? What does it take the growth of finance to have been? What would it lead a reader to expect about the recovery? And what evidence, if it appeared, would count against it? The same questions are put to the leading non-Marxist explanations, several of which describe the mechanism of the collapse more precisely than any Marxist account does, and one of which has the strongest claim of any framework to have described that mechanism in advance. Readers who finish the article should be able to say not merely that Marxists explained the crash, but which Marxist explanation they find persuasive, what it costs to hold it, and where its rivals press hardest.

One discipline governs the treatment of prediction, and it deserves stating before any claims are examined. Writers within the Marxist tradition had warned, some of them for years, that the financial expansion was fragile. Economists outside the tradition had issued warnings of comparable or greater specificity, including a paper delivered at a central banking conference three years before the collapse and a body of work from economists at the institution where central banks coordinate. Nobody, in either camp, predicted the timing: not the season, not the month, not the institution whose failure would turn a correction into a panic. The honest claim is about frameworks, not foresight. Certain ways of analysing capitalism directed attention toward the fragility; others directed attention away from it. That is a significant fact about the frameworks. It is not a prophecy fulfilled, and treating it as one corrupts the analytical content that makes the frameworks worth studying. The sections that follow keep returning to that distinction, because nearly every popular conclusion drawn from the crash depends on blurring it.

The sequence the record establishes

Begin with the expansion of mortgage credit in the United States during the first half of the decade. Lending standards for home purchase had been loosening for years, and the loosening accelerated as house prices rose: borrowers with weak credit histories, irregular incomes or no documented income at all were offered loans whose affordability depended on continued price appreciation or on refinancing before an introductory interest rate expired. Such lending was not a marginal activity conducted at the edges of the system. Contemporary accounts describe it as a substantial and growing share of new mortgages at the peak, originated by specialised lenders, mortgage brokers and the mortgage arms of major banks, and funded through capital markets rather than through the deposit base of the originating institutions. The point to hold firmly is that the loans were made to be sold. Origination had become a fee generating step in a chain, and the chain’s economics rewarded volume: each participant earned income at the moment of handing the risk to the next participant.

Securitisation supplied the machinery of the handover. Mortgages in their thousands were pooled, and claims on the pooled payments were sliced into tranches ranked by priority: the holders of senior tranches would be paid first and bore loss last, while junior tranches absorbed the first defaults in exchange for higher returns. These mortgage backed securities were themselves repackaged into further structures, collateralised debt obligations, whose tranches could receive high credit ratings even when built from the riskier slices of the underlying pools. Rating agencies assessed the structures using models whose assumptions about house prices and about the correlation of defaults across regions proved badly wrong; the historical data fed into the models came largely from a period when national house prices had not fallen in a sustained way, so the models treated a national fall as close to impossible. Investors across the world, pension funds, municipalities, banks and money market vehicles among them, bought the highly rated paper in enormous quantities, attracted by yields slightly above those on government debt carrying the same ratings. The risk had not been eliminated by this dispersal. It had been rendered illegible: nobody holding a senior tranche could easily say which neighbourhoods, which origination practices and which assumptions stood behind it.

The correction began when American house prices stopped rising in 2006 and then fell, first slowly and then across the whole national market. Borrowers whose loans had been scheduled for refinancing found the collateral worth less than the debt; defaults and foreclosures rose; the junior tranches of mortgage structures absorbed losses and the senior tranches, priced on the assumption that such losses were remote, became impossible to value and therefore impossible to sell. Institutions that had financed holdings of the paper with very short term borrowing discovered that the borrowing could not be rolled over. In August 2007 a major French bank froze redemptions in funds exposed to American mortgage assets, citing an inability to value them, and the market for asset backed commercial paper contracted sharply; the European Central Bank supplied emergency liquidity to banks within days. In Britain the following month, a mortgage lender dependent on wholesale funding suffered the first widely televised run on a British bank in generations, with depositors queuing outside branches until the government guaranteed deposits. The phrase used by officials at the time, that the turmoil was contained within one segment of lending, ceased to be credible well before the segment’s losses were fully counted.

Why did the failure of one investment bank freeze credit across the world?

Because the bank stood at the centre of a web of short term borrowing, derivatives contracts and settlement obligations, so its disorderly collapse left counterparties unable to establish what they were owed or which of their own assets remained sound. Credit froze because no institution could demonstrate quickly that it was solvent, and each hoarded liquidity against the next failure.

The decisive escalation came in the autumn of 2008. Two giant government sponsored mortgage enterprises were placed under public control in September. Days later, on 15 September, the investment bank Lehman Brothers filed for bankruptcy, with reported assets above six hundred billion dollars, after negotiations for a rescue or an orderly sale failed. The contrast with the spring, when another investment bank on the edge of collapse had been pressed into a purchase supported by the central bank, was stark: this time the institution was allowed to fail, and the failure propagated through every channel the system possessed. A large money market fund holding Lehman debt announced that its shares had fallen below the value investors expected, breaking an assumption that such funds were as safe as deposits, and withdrawals spread across the money fund industry. Insurers that had sold protection against defaults on mortgage structures faced claims they could not meet; the largest of them received an emergency loan from the central bank, initially reported at eighty five billion dollars, within days of the bankruptcy filing. Interbank lending rates spiked; corporations found commercial paper markets closed; trade finance, the unglamorous credit that moves goods between continents, contracted; and stock markets fell by magnitudes not recorded since the era of the Great Depression.

What followed in the real economy matched the violence of the financial seizure. World trade volumes fell at a pace economists compared to the early nineteen thirties: the World Trade Organization later reported a contraction in the volume of world merchandise trade of roughly twelve percent for 2009, the steepest recorded in the postwar statistical series. Industrial production dropped across the advanced economies and in many emerging ones; the American unemployment rate reached ten percent in October 2009, roughly double its level before the crisis; foreclosures ran into the millions across the decade; and output losses, measured against the path forecasters had projected before the crash, were never recovered in most economies, leaving a permanent gap between the economy that materialised and the economy that had been expected. The policy response was correspondingly vast. Central banks cut policy rates toward zero, lent against collateral they would previously have refused, and then began purchasing government bonds and mortgage securities in quantities designed to expand their balance sheets several times over, a practice that entered public vocabulary as quantitative easing. Governments recapitalised banks directly, guaranteed bank debt, insured toxic assets, expanded deposit guarantees and ran large fiscal deficits, both through stimulus measures and through the automatic effect of collapsing tax revenues and rising welfare payments. In the United States, a rescue programme authorised by the legislature reached seven hundred billion dollars in headline terms. The final act of the sequence arrived in Europe: states that had absorbed banking losses onto their own balance sheets found their borrowing costs questioned, a member of the eurozone required an international rescue in the spring of 2010, and a sovereign debt crisis within the currency union occupied the following years, transmitting the crash into a second downturn for parts of the continent.

Any explanation of the crash has to be able to narrate this sequence without embarrassment, and the sequence itself already disciplines explanation. The losses began in American housing, but the panic became general because the housing risk had been sliced, rated, leveraged and distributed through institutions whose funding was short and whose exposures were mutual. An account that speaks only of lending standards cannot explain why the losses, which were large but not unprecedented by the standards of national housing busts, threatened the entire structure. An account that speaks only of the structure cannot explain why the trigger was housing rather than something else. The four Marxist explanations and their non-Marxist rivals can be read, in part, as different decisions about which link of this chain deserves the title of cause, and the decisions carry different implications for what a recurrence would look like and what, if anything, policy could prevent one.

The instruments and the incentives that built the chain

The chain deserves a slower pass, because several explanations become plausible or implausible depending on details of the machinery. A mortgage backed security is, in conception, a simple diversification device: thousands of loans are pooled so that the idiosyncratic risk of any single borrower drowning the investor is replaced by the statistical behaviour of the pool. The conception works when defaults are largely independent events, scattered by personal misfortune across a pool whose aggregate behaviour is stable. It fails when defaults become correlated, when a common factor, falling house prices across the whole market, moves thousands of borrowers into distress simultaneously. The ratings models treated correlation as a parameter to be estimated from a benign sample period; the tranche structure then converted that estimate into a promise, since the senior tranche could only suffer if losses exceeded the cushion beneath it, a threshold the models said was remote. What investors bought, in the end, was not a portfolio of loans. It was a claim about correlation, issued by a model, stamped by an agency paid by the issuer of the security, and rarely examined by the purchaser whose own mandates often required highly rated assets and forbade the due diligence that might have complicated the purchase.

What was a collateralised debt obligation, in plain terms?

It was a security built from other securities: slices of mortgage backed bonds, many of them the riskier slices ordinary investors had declined, were pooled and sliced again into new tranches, so the top slice of the second structure could carry a high rating although the material beneath was among the weakest in the original pools.

Synthetic versions removed even the need for actual mortgages. Credit default swaps, insurance like contracts in which one party pays a premium and the other undertakes to compensate for losses on a referenced security, allowed investors to take positions on mortgage bonds they did not own, and allowed dealers to manufacture exposure far beyond the supply of underlying loans. The swaps were also the channel through which losses leapt from housing into the insurance sector: an institution that had sold protection on a large scale had, in effect, concentrated on its own balance sheet precisely the correlated risk that securitisation was supposed to disperse. Funding practices completed the circuit. Investment banks and off balance sheet vehicles held long term, illiquid structures financed by borrowing that matured in days or weeks, a mismatch tolerable only while lenders believed the assets could be sold or refinanced at short notice. When belief faltered, the assets had to be sold into a market where every other holder was selling too, prices collapsed toward whatever a distressed buyer would pay, and balance sheets that had looked adequate at model prices became inadequate at market prices within weeks. Leverage translated the fall: institutions operating with borrowed funds many times their capital could be rendered insolvent by declines in asset values that a less leveraged holder would have absorbed as a bad year.

The incentive structure ran with the machinery rather than against it. Originators earned fees at origination and bore no loss when loans defaulted, since the loans had been sold; the predictable deterioration of underwriting followed, documented in subsequent investigations that found stated income loans, loans requiring no documentation, and loans whose terms depended entirely on refinancing. Arrangers earned fees for structuring; agencies earned fees for rating and faced competitive pressure from rival agencies should their judgements prove ungenerous; executives at the major institutions were remunerated in forms tied to short term returns on equity, which leverage mechanically inflates in good years; and the final investors, or the managers acting for them, earned small increments of yield for their funds by holding paper whose risks their mandates prevented them from examining closely. None of this requires a story about individual villainy, and the serious explanations, Marxist and non-Marxist alike, generally avoid one. Each actor behaved in ways the surrounding structure rewarded. The disagreements begin when one asks why the structure existed: why so much capital sought these returns, why households were willing or compelled to carry the debt, why the expansion of credit took a property form, and why profitability, wages or demand took the shape they did in the decades before the break. Those are the questions on which the rival explanations divide.

The vindication trap, named before the evidence is graded

The claim that organises this article can now be stated in full. Call it the vindication trap: the habit of treating a crash as a verdict on an entire theoretical tradition, when the tradition in fact contains several accounts of the event that cannot all be true together. If the crash was caused by a long fall in the rate of profit in production, then it was not caused, in the same sense, by wage stagnation suppressing demand, and neither of those is identical with the claim that surplus capital had nowhere to go but property. Each account names a different primary mechanism, locates financialisation differently, and points toward divergent expectations for what follows. An event consistent with all four stories at the level of loose description discriminates among none of them. Consistency is cheap: any large catastrophe is consistent with almost any theory that predicted instability somewhere, sometime, for some reason. Discrimination is the expensive operation, and it is performed by comparing the accounts against each other and against the non-Marxist rivals, asking which mechanisms the documented sequence actually displays, which auxiliary claims each account must swallow, and which observations would embarrass it.

The trap springs on sympathetic and hostile readers alike. The sympathetic version announces that Marx is back, that the crash proved the critique of political economy, that the system’s defenders have been refuted; it then selects, usually without noticing the selection, whichever Marxist account the writer already held, and presents that account as the tradition’s verdict. The hostile version mirrors it: because the four accounts disagree, the tradition is declared incoherent, as if internal rivalry were evidence of emptiness rather than the normal condition of a living research programme; mainstream economics, after all, also produced multiple explanations of the crash, and nobody concludes from their plurality that economics as such was refuted or confirmed. Both versions avoid the only work that yields knowledge. Rival explanations of the same event, stated precisely, generate different expectations about timing, sectoral patterns, the behaviour of profits before the break, the role of household borrowing, and the shape of recovery. Where expectations differ, evidence can arbitrate. The sections below therefore state each account with its expectations attached, and the comparison is deliberately unkind to all of them: every account is asked to face the piece of the record it handles worst.

A further refinement matters for everything that follows. Explanation operates at distinguishable depths, and much confusion in the literature comes from accounts at different depths being scored as competitors when they are partly complements, or scored as complements when they are actually competitors. At the shallowest depth sit the proximate mechanisms: the instruments, the funding structures, the ratings failures and the supervisory lapses through which losses became a panic. One level down sit the enabling conditions: the supply of credit, the distribution of income, the global imbalances and the monetary stance that made the mechanisms profitable to build. At the greatest depth sit the structural drivers that Marxist accounts typically invoke: the behaviour of profitability over decades, the accumulation of capital relative to the outlets available for it, the pressure of international competition on manufacturing capacity. A claim that the crash was caused by failures of supervision and a claim that it was caused by stagnant profitability are not automatically contradictory, since supervision might have restrained a machinery whose motive force came from elsewhere; but they become rivals the moment each claims primacy, the moment each says that without its factor the crash, in something like its actual scale and form, would not have occurred. The comparison that follows grades the accounts as claims to primacy, which is how their authors mostly intend them, while noting where depth differences soften the rivalry.

The explanation comparison

The table below is the article’s findable artifact, and it is designed for use rather than decoration. Each row states one explanation, Marxist in the first four rows and non-Marxist in the remaining six. The columns force the discriminating questions: the primary cause the account identifies, its reading of the growth of finance, what it leads one to expect about recovery, and the evidence that would count against it. An account that cannot fill the final column, that no imaginable observation could embarrass, has withdrawn from explanation into affirmation, and readers should mark it accordingly whichever tradition it comes from.

Explanation	Primary cause identified	What financialisation was	What it implies about recovery	What evidence would count against it
Profitability account	Weak returns in production over decades pushed capital toward financial channels	A refuge for capital unable to earn adequate returns in industry	Recovery stays fragile while underlying profitability remains low	Evidence that profitability had durably recovered before the crash
Financialisation and underconsumption account	Stagnant wages and rising inequality suppressed consumption demand	A substitute for wage income, sustaining spending through household borrowing	Recovery limited by household debt burdens and weak pay growth	Evidence that demand held up without rapid growth in household debt
Overaccumulation and built environment account	Surplus capital lacked productive outlets and was absorbed by property	A channel diverting accumulated capital into land and buildings	Recovery slow where property overhang and debt persist	Evidence that property absorbed only a minor share of surplus capital
Long downturn account	International manufacturing overcapacity generated by intense competition	A secondary sphere expanding as manufacturing margins thinned	Stagnation continues while global overcapacity persists	Evidence that manufacturing capacity and margins were healthy before the crash
Financial instability hypothesis	Stability encouraged progressively riskier financial practices	An endogenous product of tranquil conditions rewarding leverage	Recovery requires rebuilding cautious financial practice and regulation	Evidence that risk taking did not increase as the calm period lengthened
Global savings glut account	Excess saving in export surplus economies depressed interest rates worldwide	A destination for foreign capital seeking safe dollar assets	Recovery depends on rebalancing between surplus and deficit economies	Evidence that capital inflows did not materially lower borrowing costs
Regulatory and supervisory failure account	Rules and their enforcement failed to restrain known risks	A sphere that outgrew the perimeter of effective oversight	Recovery secured by repaired supervision and capital requirements	Evidence that supervisors identified and acted on the main risks in time
Monetary policy error account	Interest rates held below appropriate levels fuelled the credit boom	A byproduct of cheap money searching for yield	Recovery follows once policy settings are corrected	Evidence that tighter money would have left the boom largely unchanged
Origination incentive account	Fee structures rewarded loan volume over loan quality	A chain in which risk was passed onward at every link	Recovery follows once incentives and liability are realigned	Evidence that originators retained risk and lending standards still collapsed
Inequality and credit account	Political and social pressure expanded credit to households whose incomes stagnated	A democratisation of borrowing that masked distributional change	Recovery hampered where indebted households cut spending	Evidence that borrowing rose fastest among households with rising incomes

Read the table by columns as much as by rows. The third column divides the field most sharply. For the profitability account, the growth of finance is a symptom: capital went where returns remained when returns in production disappointed. For the underconsumption account, finance is a mechanism of demand management, politically tolerated because it allowed living standards to be borrowed rather than earned. For the overaccumulation account, finance is a spatial and sectoral diversion, the plumbing through which surplus reached the built environment. For the long downturn account, finance is a secondary theatre, expanding in the shadow of a manufacturing problem that the financial commentary mistook for the whole story. Among the non-Marxist rows, finance is respectively a behavioural product of stability, a conduit for foreign savings, a perimeter problem, a price problem, a fee problem and a distributional sedative. These are not ten phrasings of one insight. They disagree about what would have to change for the episode to become impossible, and a reader who can state that disagreement precisely has learned more than a reader who can recite any single row with conviction.

The fourth column performs a quieter service: it exposes each account to the decade after the crash. Recoveries can be inspected. Economies where household debt remained high did experience slow, consumption shy expansions; economies carrying property overhangs worked them off at different speeds; manufacturing overcapacity did not vanish; profitability measures behaved in ways their partisans and critics still contest with the same data series. None of this settles the rivalry, because recoveries are shaped by policy reactions that no account predicted in detail, but it disciplines the rivalry by attaching each explanation to expectations that events could disappoint. The fifth column, finally, is an invitation to intellectual hygiene that readers should extend to every crisis narrative they meet, including the narratives built into news coverage and examination syllabuses: ask the narrator what observation would change the verdict. A comparison in which every row can answer is a comparison among explanations. A comparison in which some rows cannot is partly a comparison among loyalties.

How can four Marxist explanations of one crash all claim support from the same record?

Because a large event leaves a large record, and each account fastens on a different stratum of it: profit series compiled over decades, wage and debt statistics, capital flows into construction, the geography of manufacturing competition. A stratum becomes evidence for primacy only when the account also explains why the other strata took the values they did.

Where, then, does the comparison leave the general question that draws most readers, the question of what caused the global financial crisis in the sense a headline intends? The defensible summary runs as follows. The losses were generated in American housing credit, amplified by securitisation structures whose risk models failed, transmitted by leverage and short term funding through institutions bound together by derivatives and mutual distrust, and converted into a general recession when credit to households and firms seized. That narrative, which few specialists of any school dispute in outline, answers the headline at the level of mechanism. The rival explanations fight one level down, over why the system had come to depend on exactly that fragile construction: why the credit had to grow, why the housing channel carried it, why returns and wages and savings were configured as they were. Marxist accounts stake everything on that lower level. Their promise is depth; their exposure is that the mechanisms at the upper level are described better by others, and that the lower level variables moved slowly, so the accounts owe an explanation of why a slow variable produced a break in a particular autumn. Each of the four pays that debt differently, and the differences are the substance of the next four sections.

The profitability account: a long problem surfacing

The first Marxist explanation reaches back furthest. Its starting point is the claim, argued in several variants, that the rate of profit in the productive sectors of the advanced economies never durably recovered from the squeeze and decline of the late nineteen sixties and nineteen seventies. On this view, the decades before the crash were not a golden age interrupted by a financial accident. They were a long period in which capital, unable to secure its accustomed returns in industry, migrated toward activities where returns could still be manufactured: property, asset trading, lending, and the whole apparatus of financial intermediation whose expansion is labelled financialisation. Finance grew, in this account, not because financiers captured the state or because regulators slept, though both may have occurred, but because the underlying economy offered too few profitable employments for the capital seeking them. The crash, when it came, was the moment the substitute returns proved fictitious: the financial superstructure had been generating claims on a productive base that could not validate them, and the repricing of those claims was the surfacing of a problem decades deep.

The argument’s inner logic follows the classical claim that the rate of profit tends to fall as capital accumulates and the composition of capital rises, a claim examined on its own terms in the guide to the tendency of the rate of profit to fall. Writers in this family differ over measurement, as the wider debate always has, but they share a refusal to treat the financial expansion as self caused. In the presentation given by Andrew Kliman, for instance, data on American corporate profitability are read as showing no sustained restoration after the early nineteen eighties, with the consequence that growth depended increasingly on debt at every level of the system. Other writers in the same family, drawing on different series and different adjustments, reach the same shape by different routes: profitability partially recovered from its trough but remained below the levels of the postwar boom, and the gap was papered over by credit. The policy conclusion embedded in the account is bleak and distinctive. If the difficulty lies in the profitability of production itself, then repairing bank supervision, however necessary, treats the rash rather than the fever; the system will generate new fragilities around whatever channels remain open, and stagnation rather than dynamism is the default expectation for the aftermath.

What would count against the profitability explanation?

A demonstration, on measurement conventions its own proponents accept, that profitability in the productive economy had returned to strong levels well before the crash and remained there, with investment buoyant and debt ratios stable. Critics also press that profit rates in several series recovered substantially from their lowest point, and that so slow a variable struggles to date a break.

The account’s strengths are genuine and should be registered before its difficulties. It explains why financial expansion was so persistent and so general: not a fashion or a fraud but a systematic displacement of activity, visible across countries with different regulatory regimes, which a purely supervisory story struggles to match. It explains the dependence of growth on debt, the recurrent asset booms of the preceding decades in equities, property and credit instruments, as episodes of a single condition rather than unrelated accidents. And it connects the crash to the slow variables, wage pressure, productivity, international competition for manufacturing, that any complete history of the period must somehow include. Its difficulties cluster at two points. The first is measurement: profit rate series are notoriously sensitive to choices about capital valuation, depreciation, the treatment of financial profits and the boundary between productive and unproductive activity, and the literature contains careful reconstructions pointing in opposite directions, a problem assessed directly in the companion study of the rate of profit empirical evidence. The second is timing. Profitability did not collapse in 2007. By most readings it had been mediocre for a generation, through long stretches without systemic crisis. The account must therefore specify the transmission from a chronic condition to an acute event, usually by arguing that debt accumulation, itself the symptom, eventually crossed thresholds of sustainability. That move is reasonable, but it borrows the acute phase of the explanation from the credit dynamics that other accounts place at the centre, and a critic is entitled to ask how much of the explanatory work the chronic variable is still doing once the borrowing is complete.

A fair verdict on the profitability account is that it is the deepest of the four and the least self sufficient. It tells a reader why a credit fuelled expansion was likely to be attempted; it is comparatively quiet about the particular instruments, and it cannot by itself date anything. Its proponents sometimes resist this characterisation, presenting the profit rate as the master variable from which the rest follows by necessity. Readers should resist the presentation while respecting the evidence programme behind it: the account stands or falls with long run statistical reconstruction, which is slow, revisable and genuinely scientific in character, and its fate will be decided in that literature rather than in polemics about the crash itself.

The financialisation and underconsumption account: debt in place of wages

The second account begins not with capital’s returns but with labour’s income. In the decades before the crash, across most advanced economies and emphatically in the United States, wages for the majority of workers grew slowly or stagnated while productivity and top incomes rose. Orthodox growth would then face a demand problem: an economy that pays its workers a shrinking relative share must find buyers for an expanding product. The buyers, on this account, were found through borrowing. Household debt climbed to heights without peacetime precedent relative to income, carrying consumption, and notably housing consumption, that wages alone could not have carried. The state tolerated and often encouraged the substitution, since credit growth delivered rising living standards without the distributional conflict of raising pay or the fiscal cost of public provision; access to borrowing was celebrated in the language of democratised finance even as its function, in this telling, was to defer a reckoning between what the economy produced and what its population was paid. Financialisation, here, is not a refuge for surplus capital. It is the institutional form taken by a distributional settlement, and its expansion is explained by the stagnation beneath it.

The intellectual lineage runs through the monopoly capital tradition associated with the Monthly Review school and its account of surplus absorption, updated by writers such as John Bellamy Foster and Fred Magdoff, who argued in the years immediately before the collapse that the financial explosion was the system’s response to a chronic tendency toward stagnation in the underlying economy. Their formulation joined underconsumption to financialisation explicitly: as the surplus generated by production outran profitable investment outlets and mass consumption, finance ballooned to absorb it, and household borrowing ballooned to sustain demand, until the debts themselves became the crisis. The crash, on this reading, was the moment the substitution failed. Households whose real incomes had not risen could service their debts only while house prices rose and refinancing stayed open; when prices fell, the demand prop and the asset backing the loans gave way together, and the economy discovered that consumption financed by asset appreciation is consumption that vanishes with the appreciation.

Why did household borrowing grow so much while typical wages stagnated?

Because borrowing was the available bridge between the costs households were expected to sustain, housing in decent school districts, health care, education, retirement provision, and incomes that were not keeping pace. Lenders supplied the bridge eagerly, securitisation rewarded them for supplying it, and public policy preferred private debt to higher wages or expanded public services.

This account handles certain facts with notable economy. The concentration of the crash in household mortgage debt, rather than in corporate borrowing, fits its emphasis. The geography fits too: the countries with the most dramatic household debt expansions and housing booms, the United States, Britain, Ireland, Spain, suffered the deepest financial distress, while economies where household borrowing grew less tended to suffer the recession as a trade shock rather than a banking collapse. The aftermath also fits uncomfortably well: recovery in the heavily indebted economies was slow precisely along the dimension the account would predict, with households repairing balance sheets, consumption subdued, and monetary policy pushing against the constraint of private debt overhang. The account moreover assigns financialisation an explanation rather than treating it as one, answering the standing warning that the term often names the phenomenon to be explained. The growth of finance is derived from stagnation and distribution, not merely redescribed in a Latinate vocabulary.

Its difficulties are of two kinds. First, the demand shortfall it posits did not obviously bind before the crash: consumption shares of output were high, not low, in the Anglo Saxon economies, which is precisely what the debt substitution is invoked to explain, but the explanation then risks circularity, since the absence of visible underconsumption is credited to the mechanism whose breakdown is the crisis. Proponents answer that the counterfactual is the test: remove the debt growth and demand would have sagged earlier and more gently, producing stagnation rather than panic. The answer is coherent, though it moves the account’s central claim into counterfactual territory where evidence discriminates weakly. Second, the account must explain the behaviour of business investment, which was not conspicuously weak in the years before the crash in the way a simple stagnation story might suggest, and the corporate sector in several countries sat on rising cash balances, a fact that fits surplus absorption arguments better than wage led demand arguments. The two Marxist accounts thus divide the evidence between them in an instructive pattern: the profitability story is strongest on the corporate side, the underconsumption story strongest on the household side, and the temptation to merge them into one synthesis should be resisted until someone specifies how the mechanisms interact rather than merely listing both. How the process of financialisation itself is theorised, beyond this episode, is treated in the companion article on financialisation in Marxist theory, which follows the term across its several uses.

The overaccumulation account: surplus capital and the built environment

The third explanation shifts attention from flows of income to the stock of accumulated capital and its search for employment. Capitalism, on this view, periodically produces more capital than can be invested at acceptable returns in the activities that generated it, and the surplus seeks outlets wherever barriers can be lowered: foreign markets, new sectors, state expenditure, or the built environment. Property occupies a privileged place among the outlets because construction absorbs large sums slowly, because land values can be inflated by credit, and because the state can underwrite the absorption through infrastructure, planning decisions and guarantees. The resulting switch of capital into real estate defers the overaccumulation problem rather than solving it: the buildings remain, the debts incurred against expected future rents remain, and when the expected rents or resale values fail to materialise, the deferral expires in a property crash that drags the banking system down with it, since the banking system supplied the switch with its leverage. David Harvey, the account’s best known exponent, placed the crash within a career long analysis of such spatial and sectoral fixes, and presented the urbanisation of capital, in China as much as in the United States, as the great absorber of the surplus generated in the decades of expansion.

How did surplus capital end up concentrated in property?

Through a conjunction the account treats as systematic: productive opportunities in older industrial economies looked limited relative to the funds seeking placement, deregulation freed mortgage credit, governments welcomed construction and wealth effects, and property offered collateral whose apparent solidity satisfied risk models. The conjunction recurs, so property booms become the characteristic destination of surplus in mature economies.

The strengths of this reading lie in its fit with the sectoral facts and with the international pattern. The crisis was, before it was anything else, a property crisis: residential in the United States, Ireland and Spain, commercial in Britain, intertwined with construction booms whose collapse left half finished developments as the episode’s enduring photographs. The account also travels well geographically, since it predicts that the form of the crisis in each country follows the form of the preceding absorption, construction led where construction had boomed, export led where the surplus had been exported. Its treatment of the state is integral rather than appended: states do not merely rescue the system at the end; they help build the switch at the beginning, through guarantees, tax treatment, land policy and the sponsorship of home ownership as social policy. Where the account strains is in specifying the surplus. Overaccumulation is a relation between capital and profitable outlets, and demonstrating it requires the same contested profitability terrain as the first account, entered from a different direction; critics note that the concept can expand to cover any investment boom whatever, since a boom that ends badly can always be redescribed as surplus seeking an outlet. Proponents reply with the historical sequence of switches, from railways to suburbanisation to the property cycles of the nineteen eighties and two thousands, but the reader should notice that the reply is a pattern argument, powerful as narrative and weaker as measurement. The account also sits in visible tension with the underconsumption reading: one locates the difficulty in too much capital relative to outlets, the other in too little purchasing power relative to output, and while sophisticated versions attempt to derive each from the other, the plain versions predict different remedies and different fragilities, a divergence the comparison exists to expose rather than to heal.

The long downturn account: overcapacity across world manufacturing

The fourth explanation is in some respects the most austere, because it assigns finance a subordinate role and relocates the crisis in the competitive structure of world production. Associated above all with the work of Robert Brenner, whose analysis appeared a decade before the crash, it holds that from the late nineteen sixties the advanced economies entered a long downturn driven by overcapacity and overproduction in manufacturing. New entrants, first Japan and Germany rebuilding and industrialising, later the East Asian economies and eventually China, expanded capacity in industry after industry faster than demand for the output grew, while established producers, carrying heavy sunk costs and workforces they could not lightly discard, refused to exit. The result was chronic pressure on manufacturing prices, margins and profitability that no single firm could escape and no national policy could cure, since the overcapacity was international. Responses included wage restraint, currency depreciation, relocation, and sustained borrowing by states and households to sustain demand in the face of the squeeze; asset booms, in equities in the nineteen nineties and property in the two thousands, were in this telling episodes within the downturn, mechanisms by which demand was propped and losses deferred, not signs of restored health.

What separates the long downturn account from the profitability account?

The mechanism assigned to weak profitability. The classical version derives falling returns from the rising composition of capital within production, a tendency internal to accumulation. The long downturn account derives weak returns from competition between capitals internationally, too much capacity chasing too little demand. The first predicts pressure wherever accumulation proceeds; the second ties it to the geography of entry.

Read against the crash, the long downturn account explains the long prehistory: the succession of bubbles, the dependence of the American economy on consumption financed by borrowing while its manufacturing trade balance deteriorated, the recycling of export surpluses into dollar assets that held down interest rates, and the political impossibility, across administrations of different colours, of restoring the growth rates of the postwar decades. Its reading of the crash itself is deflationary in a specific sense: the financial collapse was the form the downturn took when the props failed, and a reader who fixes attention on the props mistakes the theatre for the play. The account’s vulnerabilities mirror its ambitions. Manufacturing overcapacity is itself difficult to measure at world scale, capacity utilisation data are national and uneven, and the squeeze on manufacturing sits awkwardly beside aggregate corporate profit shares in the United States that reached high levels in the years before the crash, swollen partly by financial profits whose exclusion the theory must then justify analytically. Critics within the tradition have pressed exactly this point, disputing Brenner’s reading of the post nineteen ninety five period and arguing that profitability had substantially recovered, with the crash requiring a different explanation. Brenner’s rejoinders, and the rejoinders to them, form one of the more genuinely empirical disputes in the literature, turning on series construction and sector definitions in ways an informed reader can follow. That the dispute exists, and remains unresolved after a decade and a half, is itself one of the facts this article exists to report: the tradition’s most testable account of the crash is contested inside the tradition on evidence, which is what the vindication rhetoric of the newspaper features never mentions.

The four accounts can now be laid against one another without caricature. They share a negative thesis: none accepts that the crash was a bolt from a clear sky, an accident befalling a fundamentally sound order. They share a preference for slow variables over sudden vices. Beyond that, they diverge on the identity of the slow variable, on whether finance was symptom, mechanism, destination or sideshow, on whether the binding scarcity was profitable outlets, mass purchasing power, or tolerable levels of international competition, and on what the aftermath should look like if the diagnosis is right. A reader choosing among them is choosing among research programmes with different evidentiary centres of gravity: national profit statistics, household balance sheets, the geography of construction, world manufacturing data. That is a real choice, and it is the choice the popular vindication story prevents its audience from knowing exists. The families into which these positions fall, and their longer pedigrees inside Marxist crisis theory, are mapped in the series guide to Marxist crisis theory, which this case study presupposes and applies.

The non-Marxist explanations, stated at full strength

Honest comparison requires the rivals to be more than foils. The crash generated a large mainstream literature, and several of its products explain aspects of the episode better than any Marxist account on offer. Six positions deserve full statement: the financial instability hypothesis, the global savings glut, regulatory and supervisory failure, monetary policy error, the incentive account of origination and securitisation, and the inequality and credit account developed within mainstream economics itself. Each is given here in the form its proponents would recognise, with its genuine explanatory victories acknowledged before its limits are marked.

Minsky and the financial instability hypothesis

Hyman Minsky, an American economist who worked outside both the mainstream consensus of his day and the Marxist tradition, constructed the explanation with the best retrospective claim to have described the crash’s mechanism before it occurred. His financial instability hypothesis, developed from the nineteen seventies and stated most fully in his 1986 book Stabilizing an Unstable Economy, holds that a capitalist economy with sophisticated finance is endogenously unstable: stability itself alters behaviour in the direction of fragility. In tranquil times, borrowers and lenders alike revise their sense of acceptable risk. Financing postures migrate along a spectrum Minsky named with care: from hedge finance, where cash flows cover interest and principal; to speculative finance, where cash flows cover interest but principal must be rolled over; to Ponzi finance, where cash flows cover neither and the position survives only if asset prices keep rising. Success validates risk taking, margins of safety erode, innovation stretches around regulation, and the system approaches a condition in which a modest disappointment, a rise in rates, a fall in asset prices, a failure somewhere in the chain, forces a cascade of selling and default. The moment when the reversal takes hold acquired, in commentary during the crash, the label Minsky moment, a phrase coined earlier by a fund manager and attached to a thinker most of the financial press had never read while he lived.

What is Minsky’s explanation of the crash, and how precise is it?

Minsky explains the crash as the phase of a migration toward fragile finance: years of stability taught borrowers to depend on refinancing and appreciation, taught lenders to accept thinner safety margins, and built positions that a fall in house prices converted into forced liquidation. The mechanism anticipates the sequence in detail while remaining silent about its sectoral and international shape.

The compatibility question posed about Minsky deserves a direct answer, since seminar rooms argue it with some heat. Minsky is not a Marxist, and his framework lacks the theory of value, the account of exploitation and the class analysis that define the Marxist accounts; his intellectual debts run to Keynes and to institutional economics. But compatibility is not identity, and at the level of mechanism the hypothesis interlocks with several Marxist accounts rather than colliding with them. A Marxist can accept Minsky’s dynamics as the description of how a crisis unfolds within finance, while insisting that the dynamics were set running, or given their scale and direction, by conditions in production, distribution or accumulation that Minsky leaves undertheorised. Something similar happens in reverse: economists formed by Minsky frequently accept that stagnating wages and rising debt belong in the story, while declining the Marxist theory of where stagnation originates. The boundary is therefore porous at the level of mechanism and firm at the level of foundations, and readers who ask whether Minsky supports or refutes Marx are usually confusing those levels. What Minsky does refute, cleanly, is any claim that a detailed advance description of the crash mechanism was a Marxist monopoly. The most precise anticipation in the literature came from a heterodox Keynesian writing decades earlier, and honesty about the crash begins by conceding the point.

The savings glut account, associated with the leadership of the American central bank in the years before the crash and elaborated extensively afterward, locates the cause outside the deficit countries altogether. Following the Asian financial crisis of the late nineteen nineties, several emerging economies resolved to accumulate foreign exchange reserves as self insurance, while export led growth strategies and, later, high oil prices generated vast dollar earnings in surplus economies. This wall of saving, seeking safe and liquid assets, flowed into American and European bond markets, holding long term interest rates down even as central banks raised short term ones, the conundrum a central bank chairman remarked upon in 2005. Cheap long term credit, on this account, was a global condition imposed on domestic policy: households borrowed because money was nearly free, house prices rose because credit was abundant, and the financial system’s dangerous innovations were attempts to manufacture the safe assets the world’s savers demanded, mortgage securities stamped with top ratings serving as substitutes for scarce government bonds. The account has real explanatory reach. It handles the international symmetry of the boom, the otherwise puzzling fall in long rates during a tightening cycle, and the appetite of conservative foreign institutions for paper that later proved toxic. Its weakness is symmetrical to its strength: it treats the configuration of surpluses as given, where Marxist critics ask why the world economy generated export surpluses that could only be recycled as debt in the deficit economies, and it says little about why domestic supervision permitted the manufacturing of pseudo safe assets at the scale it reached. Defenders answer that an explanation of interest rates need not also be an explanation of everything, which is fair, and which concedes that the account operates at the level of enabling conditions rather than total cause.

Regulatory and supervisory failure forms the explanation most common in official post mortems. The inquiry commissions and reviews conducted in the United States, Britain and the eurozone documented a landscape in which responsibility was fragmented among agencies, in which large parts of the system, investment banks, off balance sheet vehicles, the swaps market, money market funds, stood outside the perimeter of prudential regulation designed for deposit taking banks, and in which supervisors who possessed powers often declined to use them against the prevailing consensus that markets were managing risk well. Capital rules themselves contributed, since the international framework permitted major banks to hold thin capital against highly rated securities and rewarded the migration of assets into forms the rules measured kindly. The account scores heavily on the institutional detail: it can point to named decisions, exemptions granted, examinations not performed, warnings received and not pursued, and its remedies follow directly, broader perimeters, more capital against trading assets, supervision of systemic importance rather than of legal form. Its limits are those of all accounts centred on the referees. Supervisory failure was international, occurring under governments of different parties and regulatory philosophies, during decades when the regulated were celebrated as national champions and supervisors who restrained booms risked blame for the growth forgone. Explaining why the failure was so general requires stepping back toward the political economy the Marxist accounts inhabit, a step the official reports mostly decline to take. And even perfect supervision of the kind proposed leaves untouched the prior question of why the pressure to expand credit was so relentless in the first place.

Monetary policy error narrows the regulatory story to a single instrument. In its best known form, advanced by an economist celebrated for a rule describing how central banks ought to set interest rates, the account holds that American policy rates were held well below the rule’s prescription during the years 2002 to 2005, in response to deflation fears after the collapse of the technology stock boom, and that the deviation fed the housing expansion by cheapening mortgage credit at exactly the moment underwriting was loosening. Comparable deviations were identified in parts of the eurozone, where a single policy rate fitted to the union’s average left fast growing members with money far too cheap for their conditions, sponsoring construction booms in Ireland and Spain that ended as catastrophically as the American one. The account has the virtue of a falsifiable core and of cross country reach. Against it stand two substantial objections. Long term rates, which price mortgages, decoupled from policy rates during the crucial years, the savings glut observation returning to complicate the story; and countries that ran tighter policy by the same measures sometimes housed property booms of their own, suggesting that cheap policy money was an accelerant rather than the fuel. Critics of the account also note the retrospective character of its confidence: the deflation feared at the time was a genuine risk, and a policy judged an error after the fact was defensible with the information then in hand, which is precisely why explanation by policy error so easily becomes a search for culprits. The sober residue is that monetary settings formed part of the enabling conditions, and that no serious Marxist account denies it; the dispute is over whether the settings originated the expansion or merely failed to restrain forces arising elsewhere.

The incentive account stays closest to the machinery described earlier in this article and requires no macroeconomic theory at all. Its claim is that the originate to distribute model severed the link between lending and loss at every joint of the chain. Brokers were paid on completion, originators on volume, arrangers on structuring fees, rating agencies on mandates won, executives on short term earnings that leverage flatters, fund managers on yield gathered within mandates that treated ratings as knowledge. Under such a geometry, the deterioration of underwriting was not a lapse of diligence but the system’s rational output, and the discovery after the crash that documentation was falsified, appraisals inflated and risk models gamed reads as confirmation rather than surprise. The account explains the quality collapse of the loans better than any rival, mainstream or Marxist, and it explains why warnings from inside the industry, which existed, carried no weight against fee income. What it does not explain is the demand side of the chain: why final investors in such quantity wanted the product, why issuers could place ever weaker structures at ever finer prices, why the boom concentrated in housing rather than in some other fee rich asset. Incentives describe the slope; they do not supply the weather. Marxist accounts therefore tend to absorb the incentive story whole, as the proximate mechanism their deeper variables operated through, and the absorption is legitimate. The corresponding danger, flagged earlier, is that after absorbing every strong proximate account, a deep theory can claim compatibility with any sequence of events and forfeit the discriminating sharpness the comparison is meant to enforce.

The last of the six is the most interesting for this article’s purposes because it converges with one of the Marxist accounts from inside mainstream economics. Developed most fully in a book published in 2010 by an economist who had served as chief economist at the International Monetary Fund and written extensively on finance and development, the inequality and credit account argues that rising income concentration in the United States generated political pressure to expand access to credit as a substitute for redistribution. Let them eat credit, in the phrase the argument attracts: households whose incomes stagnated were offered borrowing instead of transfers or wage growth, officials of both parties promoted home ownership and credit access as social policy, regulators were discouraged from restraining lending that wore the aspect of inclusion, and the resulting household debt mountain made the system fragile to exactly the shock that arrived. Set beside the second Marxist account above, the resemblance is obvious: stagnant mass incomes, debt as the bridging mechanism, political toleration of the bridge, collapse when the bridge failed. The differences are equally instructive. The mainstream version locates the failure in policy choices and politics, remediable in principle by different choices, and treats the inequality itself as a product of technology and globalisation rather than of the capital relation. The Marxist version derives both the inequality and the credit response from the dynamics of accumulation and class power. Same mechanism, rival depths: the pairing is the cleanest illustration in the whole comparison that agreement about what happened does not settle disagreement about why, and that a great deal of post crash commentary mistook shared mechanism for settled explanation.

Which non-Marxist account best survives comparison with the Marxist four?

Different accounts win on different levels. The financial instability hypothesis gives the finest description of the dynamics inside finance; the savings glut best explains long term interest rates; the incentive account best explains the collapse of underwriting quality; the supervisory account best explains the perimeter failure. The Marxist four compete over the slow background conditions instead.

A compatibility map can now be drawn, and drawing it is more useful than another round of polemic. Minsky’s dynamics are compatible with all four Marxist accounts, which differ in what they think loaded the dynamics with their destructive charge. The savings glut is compatible with the long downturn account, whose surplus economies and deficit recycling it describes from the financial side, and sits in tension with any account that treats American policy as the boom’s prime mover. The supervisory and incentive accounts are compatible with everything at the level of mechanism and decisive about nothing at the level of cause, which is both their strength and their limit. The inequality and credit account is compatible with the underconsumption account almost to the point of merger, and rivalrous with it only over foundations. Genuine incompatibilities concentrate where the Marxist four face each other: over whether profitability recovered, over whether demand or accumulation was the binding constraint, over whether finance was refuge, bridge, destination or sideshow. The crash literature’s polemical energy was mostly spent across the Marxist and mainstream boundary. Its unresolved questions mostly live inside it.

The prediction question, examined without charity and without spite

No claim did more work in the public reception of the crash, and none survives examination in its popular form. The claim that Marxists predicted the crash circulated in several strengths. At its strongest, it asserted foresight: the tradition knew the collapse was coming. At a middle strength, it asserted diagnosis: Marxist analysts had identified the specific fragilities, housing debt, securitisation, stagnant wages, that in fact gave way. At its weakest, it asserted only orientation: a framework expecting recurrent crises of accumulation was unsurprised by an event that embarrassed frameworks premised on self stabilising markets. The strengths must be graded separately, because the evidence differs sharply among them, and because sliding between them mid argument is the commonest dishonesty in the literature this article surveys.

Did Marxists predict the financial crash?

Some Marxist economists warned of underlying fragility, and one influential analysis had argued a decade earlier that a long downturn was underway. Several non-Marxist economists issued comparably specific warnings. No one predicted the timing, and the tradition contains rival explanations of what happened.

Take the strongest version first, since it is the one newspaper features implied. Prediction in any serious sense requires specificity about mechanism, approximate timing and the class of event, stated in advance in a form that could have been wrong. By that standard the record is unforgiving to everyone. The long downturn analysis of 1998 argued that the world economy remained trapped in stagnation and would generate instability; it did not forecast a mortgage centred banking collapse a decade later. The Monthly Review writers described the financial explosion and the stagnation beneath it with genuine prescience about the system’s dependence on debt, in articles published as the crisis broke and in the years immediately before; the dependence was correctly identified, the detonator and the date were not specified, and could not have been from within the analysis. Warnings phrased at the level of the system will eventually coincide with an event, because systems of this kind do produce crises, and a stopped clock of this design is right whenever the crisis arrives. Foresight of timing, the thing that would have made the prediction claim remarkable, is simply absent from the record, and the article’s discipline requires saying so without softening.

The middle version fares better, and deserves its due. Several Marxist writers had indeed identified, in advance, elements that became central: the unsustainability of household debt growth relative to incomes; the transformation of banking toward fee driven origination and trading; the reliance of demand on asset appreciation; the hollowness of risk dispersion claims as leverage rose. A reader who had absorbed that literature entered 2007 less surprised than a reader formed by efficient market orthodoxy, and that difference is a real datum about the frameworks. But the datum must be compared, not merely collected. The same years produced warnings from outside the tradition that were at least as specific and in one case more so. An economist of housing markets had documented the bubble in American house prices in a 2005 edition of a well known book, with price to rent and price to income ratios laid out quantitatively. Economists at the Bank for International Settlements had spent the middle of the decade arguing, in print, that credit booms and asset price misalignments were building systemic risk and that central banks should lean against them. And a paper delivered at the Federal Reserve Bank of Kansas City’s Jackson Hole conference in 2005, by an economist who would later head a major central bank, asked whether financial development had made the world riskier and answered with a detailed account of how incentive structures in modern finance concentrated tail risk precisely in the hands of those least able to bear it. That paper, presented to the assembled leadership of world central banking in the year the mortgage boom crested, is about as specific a warning as the pre crash record contains, and it came from inside the mainstream the vindication story casts as blind.

What separates a warning of fragility from a prediction of a crash?

A warning identifies a condition that makes bad outcomes more likely: debt ratios that cannot rise forever, risk postures dependent on refinancing, prices detached from the incomes meant to validate them. A prediction commits to the outcome arriving through a stated mechanism within a horizon that lets the claim fail. Warnings can be vindicated without their authors having predicted anything.

The weakest version of the claim, orientation, is true and modest. A body of theory organised around the recurrence of crisis, the endogeneity of financial expansion and the instability of accumulation was never going to be intellectually wrong footed by a crisis in the way that a decade of triumphalist commentary about the end of boom and bust left its authors wrong footed. But orientation is cheap across many frameworks: Minsky’s followers shared it, students of Kindleberger’s historical catalogue of manias and panics shared it, and any historian of nineteenth century railway shares or nineteen twenties margin lending possessed a version of it. What the Marxist tradition distinctively contributed was not the expectation of crisis but the candidate explanations of the crisis’s roots, and those explanations, being four and incompatible, return the discussion to the comparison where it belongs. The honest formulation, to be used whenever this episode is taught or cited, is that certain analytical frameworks identified the fragility and others obscured it, that this is a claim about frameworks rather than about foresight, and that the frameworks must then submit to comparison at the level of mechanism and cause, where their real differences and real merits lie.

What Marx actually wrote about credit, and what it can bear

Claims that the crash vindicated Marx himself, rather than any living Marxist, usually rest on a picture of his crisis theory drawn from slogans. The texts support something more interesting and less convenient. Marx never published a finished theory of crisis; the relevant material lies scattered across the third volume of Capital, assembled by Engels from manuscripts, in the notebooks published as Theories of Surplus Value, and in his journalism on the commercial crises of 1847 and 1857, the latter an event he watched with intense excitement as the first genuinely world embracing crisis of the industrial system. Within the third volume’s long treatment of interest bearing capital, several ideas recur that later writers have legitimately drawn upon. Credit appears as a lever that stretches accumulation beyond its immediate limits, allowing production and speculation to expand on the basis of expected rather than realised returns. The category of fictitious capital names claims, shares, bonds, titles to future revenue, whose prices move by their own logic of capitalisation and can drift far from the value of any underlying activity, flourishing precisely because each holder can sell the claim to another before the underlying revenue is tested. And the crisis appears as the moment of forced reconciliation, when claims must be presented for payment in money and the chain of mutual credit collapses back onto the narrow base of actual settlement.

Handled with care, these materials genuinely illuminate 2008. Securitised structures were fictitious capital in an almost textbook sense: capitalised claims on future mortgage payments, priced by expected revenue streams, traded at valuations the underlying payments could not sustain, and marked to a reality only when settlement was demanded and refused. Marx’s journalism on 1857, with its attention to bills of exchange, accommodation paper and the difference between genuine commercial credit and speculative paper raised on the expectation of rising prices, reads as a recognisable ancestor of the post crash inquiries into documentation fraud. The concept of the crash as a monetary event, a sudden universal demand for means of payment in place of the credit instruments that had circulated as their equivalent, captures the liquidity panic of September and October 2008 with an economy that purely real side accounts of the crisis cannot match: for several weeks the world’s problem was not that assets were worth less but that nobody would accept anybody else’s promises in exchange for them.

Care, however, cuts the other way. The manuscripts were written against a monetary system anchored in gold, in which the scarcity of the settlement medium was a fact of metallurgy as much as of policy; the modern system of fiat money and lender of last resort capacity differs in ways that change the crisis mechanics, as the sheer scale of central bank balance sheet expansion after 2008 demonstrated. Marx’s scattered remarks do not contain a theory of securitisation, of deposit insurance, of too big to fail, or of the political management of a modern banking rescue, and his expectation that crises would deepen across successive cycles is a claim his followers have argued about rather than a finding the crash established. The respectable claim is continuity of analytical instinct: an insistence that credit is not a veil over a real economy that would otherwise proceed smoothly, but a constitutive part of accumulation whose expansions and collapses belong to the system’s normal functioning. That instinct separates the Marxist accounts from frameworks in which finance is a friction or a sideshow. It does not select among the four accounts, each of which can quote the instinct in its favour, and it does not convert a nineteenth century manuscript tradition into a forecast of twenty first century instruments. Readers who encounter the stronger claim in commentary should apply the article’s standing test and ask which of the four explanations, precisely, the quoted text is being used to support, and what the same text would say about the other three.

Why housing carried the expansion

A question the rival accounts must all answer, and answer differently, is why the credit expansion of the two thousands ran through residential property rather than through equities, commodities or industrial capacity. The brute facts are not in dispute: house prices rose across most advanced economies from the mid nineteen nineties, in several countries doubling or more in real terms; mortgage debt grew faster than incomes for a decade; construction shares of employment reached conspicuous heights in the worst hit economies; and the securities at the centre of the panic were claims on house purchase loans. Explanations of the crash that cannot say why the boom wore this shape are incomplete in a way their proponents rarely acknowledge.

The profitability account answers from the side of investment alternatives: returns in industry were too weak to absorb the funds, and property offered appreciating collateral against which credit could be safely, which is to say apparently safely, expanded. The underconsumption account answers from the side of the borrower: housing was where household demand could be debt financed at scale, since homes doubled as the asset families were culturally and fiscally encouraged to stretch for, and mortgage equity withdrawal converted the appreciation directly into spending power that wages were not providing. The overaccumulation account treats the built environment as the classic absorber, the sector whose long production periods and state involvement let it swallow surplus for years before the reckoning. The long downturn account points to the political function of the housing boom in economies whose manufacturing competitiveness had eroded: rising house prices and construction employment masked deindustrialisation in the statistics of both output and jobs. The mainstream accounts contribute indispensable pieces that none of the four should scorn: the tax and subsidy treatment of owner occupation, planning restrictions that made land scarce in successful cities, the global demand for safe dollar assets that found its way into mortgage paper, and the belief, held by borrowers and lenders in common, that national house prices did not fall, a belief with no episode in living memory against it in the United States and therefore, in the actuarial sense, no price.

The comparative moral is that housing was overdetermined, and overdetermination is a problem for adjudication among explanations. When every framework can explain the sectoral shape after the fact, the sectoral shape discriminates among none of them; the discriminating evidence must come from variables where the accounts part company in advance, profitability paths, debt dynamics, capacity measures, the sequencing of wage and price movements. This is a general lesson in the appraisal of crisis theories, worth stating plainly because syllabuses rarely do: the facts a theory was built to explain cannot subsequently be used as independent confirmation of it. Confirmation, to the extent social science admits the concept, lives in the facts the theory did not choose. Much of the post crash literature, Marxist contributions included, proceeds by exhibiting the theory’s chosen facts with mounting rhetorical confidence, and the reader equipped with the distinction has the master key to that literature’s appraisal.

The strongest objection: a failure of instruments, not of accumulation

Every sympathetic treatment owes its subject the hardest available counter case, and the counter case here is formidable. It runs as follows. The crash, on the evidence of the official inquiries, was a failure of specific financial instruments, specific ratings practices and specific supervisory arrangements, layered onto a housing correction of a size that need not have threatened the system at all. The underlying American housing losses, while severe, were of an order that a well capitalised banking system could have absorbed: subsequent official estimates placed likely losses on subprime and related mortgages in the hundreds of billions, large sums that nonetheless fall short of the capital of the global banking system whose survival was supposedly at stake. What converted absorbable losses into a systemic event was leverage, opacity and the run prone structure of shadow banking, all of them artefacts of regulatory choices made within living memory and reversible by regulatory choice. During the same decades, moreover, the slow variables the Marxist accounts invoke were also present in periods and countries that experienced no systemic crash: profitability was mediocre in the nineteen seventies in countries whose crises took entirely different forms; household debt rose for decades before it detonated; manufacturing overcapacity persisted through the long nineteen nineties expansion without a banking collapse. A cause that is present when the effect is absent, and whose movement is glacial when the effect is explosive, invites the suspicion that it is a backdrop being promoted to a protagonist. The Marxist accounts, the objection concludes, explain a financial event by reaching for production side variables that did not change sharply beforehand, because their theory instructs them to find the cause there, and the reaching is visible.

The answer available to the Marxist accounts begins by accepting the objection’s strongest plank rather than contesting it. Proximate mechanism and underlying condition are different explanatory objects, and on proximate mechanism the mainstream literature, Minsky above all, plus the inquiries into incentives and supervision, is superior to anything the Marxist tradition produced about this episode: no Marxist account of the crash describes the run on repo, the structure of the shadow banking system or the dynamics of the money fund panic with the precision of the best mainstream and official work, and pretending otherwise forfeits credibility. The Marxist claim must therefore be located where it can survive: at the level of why the system had grown dependent on the fragile mechanisms, why leverage that extreme was sought, why credit had to expand so fast to sustain ordinary growth, why the losses were located in the debts of households whose incomes had stagnated. Backdrops, the reply continues, are not causally inert merely because they move slowly. A hillside saturated over months explains the timing of a landslide badly and its occurrence well; the rainfall is the trigger, the saturation is why this hillside and not the next one gave way. The counterfactual discipline supports the distinction: supervisory regimes did vary internationally, and the crash’s incidence varied with them, as the objection says; but the pressure against which supervision was contending, the search for yield and the expansion of household and financial sector balance sheets, appeared in strict and lax jurisdictions alike, wearing the local regulatory clothes.

What survives the exchange, stated without tribal loyalty, is a division of explanatory labour with the border disputes marked. If the question is why September 2008 rather than some other month, and why these instruments and that sequence of failures, the Marxist accounts are weak and the mainstream mechanisms literature is strong. If the question is why the advanced economies had come, by 2007, to a condition in which a housing correction could detonate the credit system of the entire world, the mainstream accounts tend to fall silent or to speak of policy errors of surprisingly recent vintage for so structural a condition, and the Marxist four at least offer candidate answers, in competition with each other. Readers should also register the concession the objection extracts and keep it in view through the rest of the article: the Marxist accounts are stronger on underlying condition and weaker on proximate mechanism, and any Marxist presentation that elides the concession, dwelling on the condition as if it were the whole explanation, has failed the standard this comparison sets.

The rescues and what they disclosed

The state’s conduct during the emergency became, in popular commentary, a second theatre for vindication claims: the spectacle of governments recapitalising banks, guaranteeing liabilities and instructing central banks to purchase assets on an unprecedented scale was read by some as the practical confession that markets require the state, and by others as proof that the crisis had been a creature of state intervention from the beginning. The documentary record supports neither slogan and rewards a colder description. The interventions were improvised under extreme time pressure, inconsistent in their treatment of comparable institutions, generous in ways their architects defended as unavoidable and critics attacked as the socialisation of loss, and effective, by the narrow criterion their designers set, of preventing the collapse of the payments system and a second Great Depression of output. Shareholders in rescued institutions were sometimes diluted and sometimes spared; managements were sometimes replaced and often retained; creditors were protected with a completeness that became the central political grievance of the aftermath, since the doctrine that some institutions were too important to fail emerged from the episode strengthened rather than weakened, now carrying the authority of demonstrated practice.

For the state theory questions the episode raises, the series maintains a separate treatment in its guide to Marxist state theory and modern politics, and this article confines itself to what the rescues contribute to the comparison of crash explanations. Three observations carry weight. First, the speed and scale of the intervention falsified the picture, common in pre crash commentary, of a night watchman state retreating before global markets: the fiscal and monetary capacity deployed within weeks exceeded what most political programmes had proposed in decades, which suggests the capacity had existed and the constraint had been political rather than technical. Second, the intervention’s direction, stabilising the financial system in its existing ownership forms rather than using the moment of state leverage to restructure it, is the datum the structural Marxist accounts emphasise, reading it as evidence of the state’s embeddedness in the accumulation process it was rescuing. Third, the aftermath, in which emergency measures wound down unevenly, reform legislation enlarged supervisory powers without reversing financialisation, and the rescued sector returned to profitability ahead of the economies around it, fits none of the four accounts completely and all of them partially, a pattern by now familiar. The intervention literature thus adds a coda rather than a verdict: the state behaved as rescuer of last resort for the financial order, as several Marxist state theories would lead one to expect, and the expectation is shared with enough non-Marxist institutional analysis that citing it as a distinctive confirmation repeats, in a new arena, the original error of treating consistency as vindication.

The recovery as evidence: grading the accounts against the aftermath

Explanations of a crash are usually graded, if at all, on their narration of the fall. A sterner test grades them on the aftermath, since each account carried expectations about what recovery would require and what it would look like. The decade after 2008 supplied a recovery that was, across most advanced economies, unusually slow by postwar standards: output regained its pre crash level only after years, and in several European economies not within the decade at all; investment stayed weak relative to profits; wage growth lagged even as unemployment eventually fell; interest rates remained near zero for far longer than any forecaster had projected in 2009; and productivity growth disappointed so persistently that economists across schools reached for the vocabulary of secular stagnation, a term with a nineteen thirties pedigree that re-entered circulation through mainstream mouths. Whatever else it was, the aftermath was not a vigorous rebound that left crisis theories looking overwrought, and that fact alone distributes credit unevenly among the explanations graded here.

The underconsumption and household debt account reads the aftermath most comfortably. Economies whose households entered the crash with the heaviest debt burdens spent years deleveraging, and the drag on consumption in the United States and Britain, the slow healing of housing markets, the dependence of demand on renewed asset appreciation and on public deficits, all correspond to the account’s mechanism operating in reverse. The financialisation variant adds that the structure of finance emerged from the rescue largely intact, so the substitution of debt for wages resumed in altered forms, student debt and automobile credit expanding where mortgage credit had contracted, and the distributional condition beneath the arrangement persisted. Against the account stands the corporate evidence that troubled it before the crash: business investment’s weakness coexisted with high corporate profitability and vast cash holdings, suggesting that the constraint on spending was not a shortage of funds in the aggregate economy but a shortage of willingness to commit them, a fact that pushes explanation back toward the accumulation side.

The profitability account divides against itself on the aftermath, in a way that illuminates the whole debate. Measured profit shares in the United States recovered quickly and reached high levels, which critics of the account cite as evidence that profitability had never been the binding problem. Its proponents respond with the distinction between the profit share of income and the rate of profit on capital stock, and by pointing to the composition of the recovery, profits swollen by financial earnings, by wage compression and by the very weakness of investment that kept the capital stock growing slowly. The exchange is instructive precisely because it cannot be settled by glancing at a chart: it turns on denominator choices in capital measurement of exactly the kind the empirical literature disputes, and a reader following it learns the real state of the science, which is that the theory’s central variable is measured under conventions the theorists themselves have not standardised. Meanwhile, the long downturn account claims the aftermath as its natural habitat: a world of excess capacity, weak investment despite cheap money, competitive currency pressures and growth rates stuck below the postwar norm is the downturn continuing by other means, and the absence of a return to boom conditions a decade after the financial emergency ended is the strongest stylistic evidence in its favour. The overaccumulation account notes, correctly, that property cycles resumed their centrality with remarkable speed, that housing shortages and asset inflation returned to the politics of the same cities, and that the next expansions of credit again sought the built environment, as a theory of recurrent spatial fixes would suggest. The honest scoreboard is therefore mixed in a specific way: the aftermath vindicated the shared negative thesis of all four accounts, that no quick return to the pre crash order was available, while leaving their positive mechanisms as contested as the crash itself had. A tradition whose four rivals all predicted gloom, and were all right that gloom came, has demonstrated the fertility of its pessimism and the indeterminacy of its diagnoses, and readers should be able to hold both halves of that sentence at once.

The eurozone sequel: the crash in a second institutional form

The sovereign debt crisis that engulfed parts of the eurozone from 2010 extended the case study in a direction unusually favourable to adjudication, because the same shock ran through a different institutional architecture and produced recognisably different pathologies. Banks and sovereigns within the currency union were bound in the embrace officials later named the doom loop: states guaranteed and recapitalised their banks, weakening public finances; banks held large portfolios of their own governments’ debt, so doubts about sovereign solvency became doubts about bank solvency; and no member state possessed a central bank of its own able to act as lender of last resort in its own currency, leaving each treasury exposed to funding panics of a kind that states borrowing in currencies they issue rarely face. A rescue architecture was improvised across successive summits: bilateral loans and then dedicated funds for Greece, Ireland and Portugal, a restructuring of Greek private debt, a programme of conditionality centred on fiscal consolidation, and eventually a declaration by the union’s central bank that it would purchase member state bonds under conditions, a commitment whose announcement alone compressed borrowing costs across the periphery without a bond initially being bought. Output losses in the worst affected economies reached depths comparable to the nineteen thirties in the same countries, unemployment in Greece and Spain rose above a quarter of the labour force at the worst point, and emigration of the young reshaped the demography of whole regions.

The episode tests the explanations asymmetrically. The supervisory and perimeter accounts face the question why a currency union designed with fiscal rules and a no bailout clause generated a banking sovereign spiral more virulent than the original American crisis in several member states; their answer, that the design’s flaws were institutional and correctable, gained support when a single central bank commitment calmed the panic, suggesting that a missing lender of last resort had been a cause rather than a backdrop. The Marxist accounts emphasise what the institutional answer leaves out: the divergence within the union between a competitive export core and deficit peripheries whose industries had lost ground, the recycling of core surpluses as lending to peripheral banks and states that mirrors the global pattern of the savings glut, and the distributional form the adjustment took, in which the costs fell on public services, pensions and the unemployed of the deficit states while creditor institutions were shielded. The long downturn account, with its attention to uneven competitiveness, speaks most naturally to the core periphery structure; the underconsumption account speaks to the demand collapse inflicted by simultaneous consolidation; and the eurozone’s grim decade became, for many European writers in the tradition, the case study displacing 2008 itself. For the comparison’s purposes the sequel’s lesson is methodological: institutional design mediated how the crisis expressed itself, which no deep account may ignore, while the pattern of who lent, who borrowed and who paid followed the political economy the deep accounts describe, which no institutional account may ignore either.

Where the four accounts part company on policy

The rivalry among the Marxist explanations stops being academic when remedies are attached, because each diagnosis prescribes a different politics and each politics would, if enacted, falsify its rivals’ expectations in a different way. The profitability account implies that measures confined to finance rearrange the symptoms: capital requirements, resolution regimes and conduct rules may prevent the same failure mode, but if returns in production remain the underlying difficulty, fragility migrates to whatever sector the rules cover least, and the realistic horizon is stagnation punctuated by instability rather than restored dynamism. The underconsumption account implies a distributional remedy: sustained recovery requires mass purchasing power rebuilt through wages, public provision or transfers, and debt written down where it cannot be serviced from stagnant incomes; absent that repair, demand management through credit repeats the pre crash pattern under new labels. The overaccumulation account implies scepticism toward property led recoveries in particular and directs attention to the absorption question, what social purposes the surplus could serve if its private destinations keep failing, which is why its exponents so often write about urbanisation, infrastructure and the environment as the terrains on which the next fix, benign or catastrophic, will be attempted. The long downturn account implies that national remedies are structurally limited, since the overcapacity is world scale: one country’s devaluation exports the squeeze to its competitors, one country’s consolidation deepens it, and relief awaits either a destruction of capacity on the scale history has usually delivered through slump and conflict, or the opening of new industries capable of absorbing capital and labour at the required magnitude.

Set against these, the mainstream remedies display an answering spread. The supervisory account prescribes the perimeter and capital reforms that most jurisdictions in fact enacted in diluted form. The monetary account prescribes rules based policy to avoid feeding booms. The incentive account prescribes liability: risk retention by originators, deferred remuneration, personal exposure for decision makers. The inequality and credit account prescribes redistribution as financial policy, treating the distribution of income as a macroprudential variable. The pattern to notice is that remedy talk exposes depth commitments that cause talk can hide. Two writers can nod together through a narrative of 2008 and discover, at the moment of prescription, that one believes the episode was a correctable malfunction and the other a systemic disclosure; examinations, editorial pages and party programmes are mostly written at the level of prescription, which is why the apparently scholastic rivalry among explanations decides practical arguments far from the seminar room.

The crash also revived, in the policy sphere, the oldest comparison in this territory, between demand management in the Keynesian style and the Marxist insistence that crisis originates in accumulation itself. Emergency practice in 2008 and 2009 was improvised Keynesianism on a grand scale, deficits, monetary expansion, public guarantees, executed by governments with no doctrinal commitment to it; the consolidation that followed in much of Europe from 2010 was defended in the language of confidence and opposed in the language of demand, and the argument between those languages consumed the subsequent decade of macroeconomic debate. The companion article on Marx and Keynes on crisis follows that comparison in detail; here the relevant observation is narrower. Marxist commentators divided over the emergency measures along the lines of their crash accounts: those who read the crisis as a demand and debt event could support deficits as mitigation while doubting their sufficiency; those who read it as a profitability event tended to see state borrowing as the displacement of the reckoning onto public balance sheets, a prediction the eurozone sequel appeared to illustrate. The Keynesian mainstream, for its part, divided over whether the episode had been a liquidity panic amenable to its standard tools or a balance sheet recession demanding new ones. Crisis theory and crisis policy illuminate each other best when the mapping between them is drawn explicitly, account by account, rather than asserted for traditions in the lump.

How a historian weighs a narrative written while the buildings were still falling

The crash generated its literature in real time, and the real time literature poses source problems that the distance of even a few years makes soluble but not simple. Books and articles written during the panic of 2008 and the policy storms of 2009 possess the texture of the moment, the genuine uncertainty about whether the payments system would survive the month, that retrospective writing cannot honestly reproduce. They also carry the moment’s distortions: magnitudes guessed under pressure, institutions blamed or absolved on incomplete records, and causal stories fitted to data that would later be revised so substantially that some early conclusions rest on figures no statistical agency would now defend. The official inquiries that followed, with subpoena powers and access to internal correspondence, corrected the record on underwriting, on supervisory knowledge and on the sequence of decisions in the crisis weekends, and occasionally reversed the verdicts journalism had reached in days. A historian weighing the competing explanations therefore handles at least three strata of source: the contemporaneous narrative, valuable for perception and treacherous for fact; the inquiry stratum, stronger on documented sequence and weaker on the structural variables that inquiries are not constituted to measure; and the statistical reconstruction stratum, the national accounts, profit series and debt statistics, revised repeatedly and carrying their own conventions into every conclusion built on them.

Three working rules follow. First, separate the strata when an explanation cites its evidence: an account resting on inquiry documents about the panic weeks and an account resting on forty year profit series are not answering the same question, and scoring one against the other’s home ground produces the appearance of refutation without the substance. Second, privilege what a source was in a position to know. A central bank’s internal minutes are excellent evidence of what officials believed and poor evidence of whether their beliefs were right; a memoir by a rescued executive is evidence about the experience of rescue and a negotiating document in the politics of blame. Third, treat every account’s silences as data. The early mainstream narratives barely mention wages; the early Marxist narratives barely mention the repo market; the early journalistic narratives barely mention anything outside the United States until the eurozone forced the omission closed. Silences of that kind map the frameworks’ blind spots more reliably than their arguments map their strengths, and a comparison that reads for silence as well as for assertion will usually discover that the rival explanations were, among other things, looking at different countries, different decades and different balance sheets while believing they were describing one event. The crash was one event in the way a war is one event: a connected field of many actions, in which a narrative’s choice of theatre already contains half its conclusions.

What the crash changed inside Marxist economics

The episode’s effects on the tradition that claimed it deserve the same unsentimental handling as its effects on everyone else. Enrolments, sales and citations responded as the newspapers said: courses on Marxist political economy filled, publishers reissued classics, journals commissioned special issues, and a generation of students encountered the tradition first as an interpretive key to the unfolding crisis. Inside the scholarship, three changes stand out from the publicity. The first was a migration of attention toward finance as an object of Marxist theorising in its own right rather than as a derivative sphere: work on financialisation, on the financial expropriation of households through fees, interest and payment systems, and on shareholder value as a form of class discipline within the firm, moved from the margins of the field toward its centre, and stayed there. The second was a sharpening of the empirical disputes: the availability of longer and better profit and debt series, and the obligation the crash imposed to say something definite, pushed the profitability camps into more explicit statements of measurement convention, which is a precondition of the debate ever being settled rather than merely continued. The third was a generational change in the questions asked: younger scholars formed by the episode tended to treat crisis as a normal object of study within a functioning if unstable order, rather than as a harbinger of systemic terminus, and the tradition’s centre of gravity shifted accordingly, from breakdown theory toward the analysis of instability, distribution and state management.

Counterweights belong in the same paragraph. The crash did not produce convergence: the four accounts examined here emerged from the decade as distinct as they entered it, and their exchanges grew no gentler for the shared applause. It did not produce institutional transformation within economics as a discipline, where the commanding heights of the journals and the policy institutions absorbed the episode largely as a call for better models of the financial sector rather than for different foundations. And it did not settle the prediction question in the tradition’s favour among the wider public, where the memory that commentators had disagreed about everything except the wickedness of bankers faded faster than the memory that someone had been proved right. A clear eyed tradition should prefer the accurate inventory to the flattering one, since the inventory contains the actual assets: a set of rival, evidence engaged research programmes, a revived literature on finance and class, and a cohort trained by the episode. Assets of that kind compound quietly. They do not require the crash to have been a coronation.

The financial crash beyond the United States: three national variations

The American sequence dominates the literature for good reason, since the instruments were American and the detonator was pulled there, but the crisis wrote markedly different national stories elsewhere, and the variations supply comparative evidence the explanations compete to absorb. Iceland offers the limiting case of financialisation outrunning a state: an island economy of a few hundred thousand people whose three main banks had grown balance sheets approaching ten times national output, funded in foreign currencies and in the deposits of foreign savers attracted by high rates. When wholesale funding closed in the autumn of 2008, all three failed within a week, the currency collapsed, capital controls were imposed and remained for years, an international rescue followed, and the country conducted an unusually public reckoning, with a truth commission report, prosecutions of bankers, and two referendums in which voters rejected terms for compensating foreign depositors. No Marxist account built from American profit rates or American wages describes Iceland without translation work, since the island’s pre crash economy was booming by conventional measures; the savings glut and perimeter accounts, by contrast, translate easily, a reminder that the deep explanations purchase their American fit at a price in portability.

Britain presented a financial centre variant: a large internationally active banking sector relative to the economy, a housing boom and household debt expansion rivalling the American, and a crisis whose iconic images were a depositor run in 2007 and the part nationalisation of major banks in the weeks after Lehman. Ireland presented a property and fiscal variant: a construction and land boom of extraordinary relative size, banks whose lending was concentrated in development finance, a blanket state guarantee of bank liabilities issued in the crisis’s worst fortnight whose cost eventually reached a share of national income that forced the state itself into an international rescue programme, and an aftermath of emigration and public sector contraction. Spain, whose banks had weathered the first phase under a supervisory regime praised for its provisioning rules, succumbed in the second phase through its regional savings banks and a construction collapse that carried unemployment above a quarter of the workforce. Germany, with stagnant household debt and no housing boom, experienced the crash as an export collapse and a set of embarrassing exposures in its public sector banks, which had bought the American instruments in their search for yield. The pattern across the five is the comparison’s friend: where the preceding decade had built a property and credit boom, the crisis was domestic and banking centred; where it had not, the crisis arrived by trade and by portfolios. Any theory of the crash should be able to state, country by country, which of its variables were present, and the theories divide instructively in how gracefully they perform the exercise: the household debt and property accounts perform it almost mechanically, the supervisory account performs it through regulatory variation, the profitability and long downturn accounts perform it through the slower variables of competitiveness and accumulation, at the cost of more auxiliary assumptions per country.

Misreadings the comparison exists to correct

Certain errors recur so predictably in coverage of this episode that correcting them belongs in the body of the article rather than in its margins. The first is the singular definite article: the Marxist explanation of the crash, as if the tradition spoke with one voice, when the documentary record shows four accounts in open rivalry, a correction that this article has pursued at length. The second is the reduction to greed. Human acquisitiveness is a constant of the species and of every decade on record, including decades of financial quiet; a variable that does not vary cannot explain an event that does, and the interesting question is always why acquisitiveness was institutionally licensed, rewarded and leveraged in this period and restrained in others, which is a question about structures of incentive and accumulation rather than about the state of anyone’s soul. The third is the reduction to deregulation uttered as a complete sentence. Rules were indeed repealed, exemptions granted and perimeters left undrawn, and the supervisory account in this article records the fact with respect; but deregulation names a permission, not a pressure, and an explanation that stops at permission cannot say why the pressure to expand credit existed, why it took a property form, or why jurisdictions with different rulebooks succumbed in pattern. The fourth error is the mirror of the third: treating financialisation as an explanation in itself, when the word names the growth of finance whose causes are the very thing in dispute among the profitability, underconsumption, overaccumulation and stagnation readings. The fifth is the claim that the crash confirmed the falling rate of profit, stated as if the event itself were the evidence. Confirmation runs through the profit rate series, which antedate the crash and will outlast the arguments about it; the crash is consistent with the thesis, and consistency, as the vindication trap teaches, is the beginning of an inquiry rather than its conclusion. The sixth error, the conflation of warnings with predictions, has been given its own section above because it does the most damage: it converts a defensible claim about analytical orientation into an indefensible claim about foresight, invites the equally indefensible rebuttal once the foresight claim collapses, and buries the genuine achievement, which is that several frameworks taught their readers to watch the balance sheets that mattered while the official conversation watched the wrong indicators in tranquillity.

An older argument underneath: finance capital and its critics

The four accounts did not invent their disagreements in 2008; they inherited them from a century of argument about the place of finance in capitalist development, and a reader who knows the inheritance can predict each account’s crash narrative from its intellectual ancestry with uncomfortable accuracy. At the tradition’s origin stands the analysis of finance capital published by Rudolf Hilferding in 1910, which described the fusion of banking and industrial capital into concentrated blocs, the growing dominance of the great banks over industry, and the displacement of competitive markets by administered ones. Hilferding’s finance was powerful, organised and patient, the commanding height of a maturing system; Lenin’s wartime pamphlet on imperialism carried a version of the picture to a vast audience, and the image of monopoly finance as the system’s final stage entered the tradition’s bloodstream. Against that inheritance, the twentieth century’s later Marxist economists built in several directions at once. The monopoly capital school shifted attention to the giant corporation and its surplus, treating finance largely as an absorber. The theorists of the falling rate of profit returned to the manuscripts of the third volume and rebuilt the crisis argument on production. The regulation and social structure schools, influential in the nineteen eighties and nineties, periodised capitalism into accumulation regimes, which disposed their readers to see financialisation as the organising principle of a new regime rather than as a symptom of an old one’s distress, a disposition visible in crash accounts that treat the financial order as a system with its own coherence and beneficiaries rather than as a refuge or a bridge.

The crash stress tested each inheritance. Hilferding’s banks, patient and dominant, bore little resemblance to the frantic, short funded, originate and distribute machinery that actually failed, and accounts formed by his picture had to be substantially rebuilt before they could describe institutions that manufactured risk in order to sell it rather than holding risk in order to govern industry. The falling rate tradition faced the timing problem in its purest form. The regime theorists faced the opposite embarrassment: a framework tuned to the coherence of the financialised order had to explain why the order’s own reproduction requirements, household borrowing above all, generated its collapse, which the underconsumption accounts, regime theory’s close cousins, proceeded to do with some success. None of this intellectual history is decoration. It explains why the four accounts were available so quickly when the crash arrived, most of them within months, fully formed: the writers were not improvising explanations for an unprecedented event but deploying frameworks constructed over decades for other purposes, which is also why the frameworks disagree. A tradition that had spent a century arguing about finance was never going to stop arguing because finance obliged it with a crisis, and the speed of the Marxist response, so often cited as evidence of the tradition’s vitality, is equally evidence of its pre existing divisions, the two facts being the same fact viewed from different distances.

Quantitative easing and the distributional argument

The unconventional monetary policies that followed the rescues generated a second round of explanatory dispute, smaller than the first but instructive in the same pattern. Central banks, having cut policy rates to their effective floor, purchased government bonds and, in the American case, mortgage backed securities, in volumes that multiplied their balance sheets; the stated purposes were to lower long term borrowing costs, to support asset prices and thereby demand, and to signal commitment against deflation. The policies’ distributional incidence became a public controversy: asset purchases raise the prices of assets, assets are held disproportionately by the wealthy, and a recovery engineered through portfolios will therefore widen wealth inequality even where it succeeds in its stated aims, a mechanism critics within and beyond the Marxist tradition pressed with force. Central banks replied with their own studies arguing that the employment effects of the policies benefited lower income households most, since unemployment is the great unequaliser, and that the counterfactual of deeper slump would have distributed its harms more brutally still. The empirical literature that accumulated found elements for both: measurable wealth effects at the top, measurable labour market effects at the bottom, magnitudes contested.

For the comparison’s purposes, the episode matters because each crash account recruited the monetary aftermath as supporting evidence in its own image, repeating the original manoeuvre at one remove. Profitability writers noted that years of near zero rates and asset support failed to restore investment dynamism, reading the failure as the signature of the underlying condition their theory names. Underconsumption writers noted the dependence of demand on asset prices that policy was now obliged to manufacture, reading the obligation as their substitution mechanism promoted from private improvisation to public doctrine. Long downturn writers noted the competitive currency anxieties and the export of stagnation between blocs. The monetary authorities’ own account, that the policies worked as designed against a deflationary emergency and that structural questions lie outside a central bank’s remit, is also coherent, and highlights, once more, the division of labour between levels of explanation that this article has used throughout. The methodological point deserves underlining as the case study closes: when every subsequent development can be absorbed as confirmation by every framework, the frameworks are no longer being tested, and the reader’s task shifts from adjudication among theories to vigilance about the moment adjudication stopped. That moment, in the crash literature, can usually be dated to the point where a writer’s account of 2008 became an account of everything since, requiring no new evidence and risking no new embarrassment.

The run that had no queues: shadow banking under stress

Public memory of bank runs carries the imagery of depositors at counters, and the crash of 2008 produced that image exactly once in Britain. Its decisive runs were invisible, conducted by institutions against institutions in markets most citizens never knew existed, and any account of the episode that omits them misdescribes how close the system came to disintegration. The shadow banking system, the name given to the network of investment banks, conduits, structured vehicles, money market funds and securities lenders that performed bank like functions outside bank regulation, funded itself substantially in the repurchase agreement market, where securities serve as collateral for loans of a day or a week. In calm times the arrangement was treated as nearly riskless, since the lender held marketable collateral against the cash. The crisis revealed the arrangement’s dependence on the collateral’s price and on haircuts, the discount at which collateral was accepted: as mortgage related securities became suspect, haircuts rose, in some categories from negligible levels toward a third or more of value, which is arithmetically equivalent to a withdrawal of funding on a corresponding scale. Institutions that had appeared liquid discovered that their funding had been a confidence phenomenon; vehicles were forced back onto the balance sheets of sponsoring banks; money market funds, holding the short paper of the same complex, faced redemptions that threatened their promise of stable value; and the panic acquired the self feeding geometry of the classic run, in which each participant’s rational withdrawal deepens the insolvency it fears.

The theoretical significance of the run is that it disciplines both camps of the explanation debate. For mainstream writers formed by the deposit insurance era, it demonstrated that the institutional reforms of the nineteen thirties had not abolished runs but displaced them into the uninsured wholesale markets, and the economist who did most to formalise that lesson in the crash’s aftermath wrote of a run on the shadow banking system as the event’s core, a formulation now standard in the textbooks. For Marxist writers, the run vindicated a different emphasis: Marx’s analysis of the credit system had stressed exactly this moment, the sudden general demand for money as the only acceptable means of payment, when the instruments that had circulated as capital reveal themselves as claims whose value depends on a confidence the crisis withdraws. The convergence is real and should be recorded as such; two intellectual traditions, starting from different foundations, arrived at compatible descriptions of the panic’s mechanics, one through institutional economics and one through a nineteenth century theory of credit. Convergence at the level of mechanics, however, returns the argument to causes with the same force as before: the run explains how the collapse propagated, not why the system had organised its funding so that a collapse could propagate through it, and on that question the four accounts and their mainstream rivals resume their disagreement undisturbed. A reader who understands the repo run possesses the episode’s mechanism; a reader who also understands why the mechanism was built possesses the beginnings of its explanation.

Starting points: why some narratives open with a loan and others with a ratio

Every account of the crash chooses a first scene, and the choice is diagnostic. Narratives in the journalistic and official traditions open with a loan: a borrower in California or Nevada signing for a sum the paperwork could not support, a broker completing a file, an underwriter waiving a check. From that scene the story ascends through the chain of securitisation to the panic, and causation is located in the deterioration of practice at the chain’s origin. Narratives in the Marxist traditions open elsewhere: with a ratio in a statistical appendix, profitability across four decades, the wage share across three, the share of construction in output, the capacity utilisation of world steel. From that scene the story descends toward the loan, which appears as the downstream expression of the ratio’s long movement, and causation is located in whatever drove the ratio. Neither opening is a stylistic preference. Each encodes the writer’s theory of where explanation lives, in conduct or in structure, in events that can be filmed or in magnitudes that can only be tabulated, and each opening forecloses certain questions: the loan first narrative rarely asks why millions of such loans became signable in the same half decade, and the ratio first narrative rarely lingers on the fraud in the file.

Recognising the device improves every reader’s grip on the literature, including the literature assigned in courses, where the set texts often divide by opening scene more reliably than by declared allegiance. It also explains a persistent feature of the debate, the feeling participants report of talking past one another. A conduct narrative can be complete on its own terms, accurate in every particular, and still leave the structural reader unsatisfied, because the structural question, what configuration of the economy made that conduct profitable at that scale at that time, has not been addressed but only postponed. A structural narrative can be equally complete and leave the institutional reader unsatisfied, because the transmission from a forty year trend to a specific Tuesday in September runs through mechanisms the narrative treats as plumbing. The comparison practised in this article is, in effect, a refusal to let either completeness stand as the whole: the standard set for a satisfactory account is that it must narrate the loan and the ratio in one story, assigning each its level, without allowing the level it prefers to swallow the other. By that standard, the strongest single works in the crash literature are those, from several traditions, that begin in one scene and earn their way to the other, and the weakest are those that mistake their opening tableau for a refutation of everyone else’s.

Transmission: how a mortgage crisis became a world recession

The adjective global in the crash’s usual name is earned by its transmission, and the channels of transmission constitute another body of evidence over which the explanations contend. Four channels did the heavy work. Trade was the fastest: orders for manufactured goods collapsed across continents within a single quarter, supply chains transmitted the stoppage from final demand in the deficit economies to factories in the surplus ones, and economies with sound banks and no housing booms, Germany and Japan prominent among them, suffered contractions as deep as the crisis’s originators through this channel alone. Finance was the second: banks in Europe held the American securities and funded them in dollars, so the dollar shortage of the panic was international from its first week, relieved only when the American central bank extended swap lines that made it, in effect, a lender of last resort to the world’s banking systems. Remittances, commodity prices and capital flows formed the third and fourth channels for the developing world: migrant earnings fell, oil and mineral prices collapsed from their peaks and then partially recovered, private capital fled emerging markets indiscriminately before returning selectively, and several economies in eastern Europe that had borrowed in foreign currencies to finance their own property booms required international rescues on terms resembling the eurozone’s later ordeals. The policy responses beyond the Atlantic world also mattered to the aftermath: the largest emerging economy answered with a credit and infrastructure stimulus of enormous scale, which sustained world demand for commodities and machinery through the worst years and, in the judgement of many historians, marks the moment the centre of gravity in global demand visibly shifted.

The Marxist accounts claim the transmission pattern with varying justice. The long downturn framework, built on world scale competition from the start, accommodates the global synchronisation most naturally, since its variables were never national. The overaccumulation account points to the Chinese stimulus as a spatial fix of textbook form, surplus absorbed by construction and infrastructure on a scale dwarfing the American property boom, with the same deferred questions about the debts incurred and the returns expected. The underconsumption account notes that the deficit economies’ household retrenchment was the hole into which world demand fell, and that surplus economies’ reluctance to expand domestic consumption in compensation prolonged the shortfall, a distributional reading of the global imbalance. The profitability account treats the synchronisation as evidence that the condition was never a national policy error. Against the set, the mainstream global accounts, the savings glut above all, claim the same pattern from the flow side, and the reader faces a by now familiar configuration: a documented pattern, several frameworks that accommodate it, and discrimination possible only where the frameworks’ expectations about the pattern’s details diverge, in the sequencing of recoveries, the behaviour of investment in the surplus economies, and the fate of the debts the stimulus created. Filing the pattern as confirmation of whichever framework one brought to it is the vindication trap operating at world scale, and the size of the stage does not improve the logic.

What would change the verdict: a closing standard for the comparison

Comparisons should end by stating their own exposure. The grading offered in this article, four Marxist accounts strong on background conditions and weak on mechanism, mainstream accounts strong on mechanism and uneven on causes, the prediction claim reduced to a claim about orientation, rests on the documentary record as it stands and could be revised by identifiable developments. The profitability dispute could be narrowed toward settlement if the measurement conventions of the rival camps converged, as occasionally happens when a profession tires of a stalemate; a settled series showing durable restoration of profitability before 2007 would damage the first and fourth accounts severely, while a settled series showing stagnation would damage their critics. The underconsumption account could be tested more sharply than it has been against the investment behaviour of the corporate sector, where its explanations remain thinner than its household evidence. The overaccumulation account could acquire the measurement discipline it lacks if its exponents specified the surplus independently of the booms it is invoked to explain. Minsky’s framework could lose its privileged position in the comparison if a demonstrated case emerged of a long financial calm that ended without the migration toward fragility, though the historical catalogue makes the prospect remote. And the whole architecture of graded rivalry could be superseded if a synthesis appeared that specified the interaction of profitability, distribution and credit dynamics in a single model with discriminating predictions; several research programmes gesture at such a synthesis, none has yet commanded assent across the camps, and the gesture itself testifies that the field’s practitioners feel the force of the standard this article has applied.

Until such developments arrive, the case study’s yield for a reader is a set of portable competences rather than a crowned winner. The reader can narrate the documented sequence without borrowing any theory’s vocabulary. The reader can state four Marxist explanations and six non-Marxist ones, identify the level at which each operates, and name the evidence that would embarrass each. The reader can detect the vindication trap in its sympathetic and hostile forms, in commentary on this crash and, by transfer, in commentary on any future one, since the trap is a rhetorical structure and not a property of the episode. And the reader can hold the prediction question at its honest strength: frameworks that taught attention to fragility earned credit as instruments of orientation, nobody foresaw the autumn of 2008, and the difference between those propositions is the difference between scholarship and legend. A tradition that submits to this treatment emerges neither vindicated nor refuted but legible, its internal arguments visible as arguments, its evidence programmes identifiable, its claims priced. Legibility of that kind is the precondition of every more dramatic verdict anyone might wish to reach, and reaching it has been the purpose of the comparison from its first table onward.

A worked comparison: five shared facts through four lenses

Abstractions about levels and depths earn their keep only when applied, so consider five facts that every party to the debate accepts, and watch what each Marxist account does with them. The facts: American house prices fell by roughly a quarter from their peak on the widely used national indices, more in the boom regions; household debt in the United States reached a peak relative to disposable income in 2007 that no earlier peacetime year had approached; the American manufacturing trade deficit had widened for a decade before the crash while manufacturing employment contracted; measured corporate profit shares recovered from their early two thousands dip to high levels by the middle of the decade; and the panic phase, when it came, was a run on wholesale funding markets rather than on insured deposits.

The profitability account takes the fourth fact as its problem and the first as its occasion. Profit shares, it insists, are not profit rates, and the recovery of the share alongside weak investment and rising indebtedness is precisely the configuration its theory predicts when returns on the growing capital stock disappoint: income is captured, but not validated by an expansion of productive capacity proportionate to the claims accumulating against it. The house price fall then matters as the moment the claims were audited. The underconsumption account takes the second fact as its centre and the fourth as a corroboration: high profit shares and stagnant wages are two faces of one distribution, and the debt peak records the mechanism by which demand was sustained across the gap, until the asset that collateralised the mechanism fell in price and the mechanism reversed. The overaccumulation account takes the first fact and the construction boom behind it as its centre: the surplus had been parked in the built environment, the fall in prices is the devaluation of that parked capital, and the other facts arrange themselves as conditions and consequences. Brenner’s account takes the third fact as its centre and treats the second as the American form of demand maintenance in an economy losing industrial ground: borrowing substituted for competitiveness, and the crash arrived when the substitution met its limit. The fifth fact, the wholesale run, is handled by all four as mechanism rather than cause, which is either disciplined modesty or a shared evasion, depending on how much of the explanation a reader believes mechanism to be.

Two conclusions emerge from the exercise. The first is that the accounts are genuinely rivals and not partitions of one truth, because each demotes another’s central fact to a derivative status, and the demotions cannot all be correct: household debt cannot be simultaneously the prime mover and a mere symptom of profitability, except under a specified interaction that none of the plain versions supplies. The second is that adjudication is possible in principle because the demotions have empirical consequences. If household debt was prime, economies with similar profitability paths but restrained household credit should have escaped the worst, and the cross country pattern leans that way. If profitability was prime, the post crash investment weakness should track capital stock returns more closely than household balance sheet repair, and the evidence there is genuinely mixed. If overcapacity was prime, the recovery of manufacturing margins through capacity destruction should eventually have released a stronger investment cycle, whose arrival the long aftermath kept postponing. A reader who follows the five facts through the four lenses has done the essential work of the case study; the remaining literature is commentary on an exercise of this shape, conducted at greater length and with better data.

The insurer at the centre: concentration disguised as dispersion

One institution’s failure mode concentrates the episode’s lessons so effectively that it merits separate treatment in any full account. A global insurance group, among the largest in the world, had built through a financial products division a vast book of credit default swaps, including protection sold on the senior tranches of mortgage structures, the very instruments rated as near riskless. The division’s contracts were written on the assumption that the underlying structures would not suffer losses reaching the protected layers, so the protection was priced thin and reserved against lightly; the parent company’s high credit rating allowed the contracts to be written without posting the collateral that a lesser rated seller would have faced. When the mortgage structures deteriorated, the contracts’ market values moved against the seller on a scale that triggered collateral demands precisely calibrated to the deterioration, and the group’s liquidity, not merely its solvency on optimistic valuations, failed within days in September 2008. The central bank’s emergency loan, extended to prevent the group’s disorderly collapse, was justified publicly by the density of its connections: it insured, financed or guaranteed pieces of nearly every major institution, so its failure threatened to transmit the mortgage losses, at the worst possible moment, to the entire network at once.

The case disciplines a slogan. Securitisation was supposed to disperse risk to those most able to bear it; the insurance case shows risk being gathered, through the derivatives written on the dispersed securities, into a single node whose ability to bear it rested on a rating and a model. Dispersion at one layer of the system had been reconcentration at another, invisible to participants who monitored the layer they traded in. Each explanation claims the case at its level. The incentive account observes that the division’s managers were paid on premium income whose risks matured beyond their horizons. The supervisory account observes that an insurance holding stood outside the consolidated supervision applied to banks while conducting bank dwarfing risks. Minsky’s readers observe a hedge posture migrating to a Ponzi posture inside a single balance sheet as collateral calls converted paper losses into cash demands. The Marxist accounts observe the fictitious capital circuit completing itself: claims on claims, capitalised expectations stacked until the stack’s own weight required rescue by the state, whose intervention then preserved the stack’s owners in the name of the payments system. All four observations are true, and their compatibility is the case’s final lesson. Catastrophes of this complexity do not have a level at which the true explanation resides and other levels at which only commentary resides; they have mechanisms that interlock, and the rivalry among explanations is a rivalry over which interlocking mechanism deserves the causal emphasis, conducted, if it is conducted well, with the interlocking kept in view.

Magnitudes and their limits: what the numbers can and cannot settle

Crisis arguments are often conducted as if the right statistic, produced at the right moment, would end them, and part of this article’s purpose is to explain why that hope keeps failing. Consider the magnitudes most often deployed. Estimates of total losses on American mortgage related securities varied, across the crisis years, by hundreds of billions of dollars depending on the assumed path of house prices and defaults; official sector estimates published in the spring of 2008 were revised upward repeatedly as prices fell, which meant that every argument premised on the losses being absorbable, or unabsorbable, was premised on a moving quantity. Estimates of the output cost of the crisis vary more extravagantly still, because they require a counterfactual path for economies that never existed; studies by official institutions have placed the permanent loss of output in the advanced economies in ranges wide enough to accommodate almost any rhetoric. The fiscal costs of rescues, often cited in headline form, shrink dramatically when recoveries on the assets acquired are netted, in some countries toward small fractions of the announced figures, while the indirect fiscal costs through lost revenue and higher debt interest dwarf the direct ones and are rarely attributed with the same confidence. None of this means the numbers are useless. It means their error bands are part of the evidence, and an explanation that survives only at one end of every band is weaker than its presentation suggests.

The deeper measurements resist settlement for structural reasons the article has met repeatedly. Profit rates depend on capital stock valuation under conventions that can move the result across the very threshold the argument turns on. Household debt ratios depend on income definitions and on whether the relevant burden is the stock of debt, its service cost at prevailing rates, or its distribution across households whose circumstances differ wildly within the same average. Capacity measures in world manufacturing rest on national statistics of uneven quality compiled for other purposes. A mature reader therefore treats the statistical disputes inside the Marxist literature not as a scandal to be exposed or a pedantry to be excused, but as the actual terrain on which parts of the rivalry will be decided, slowly, by reconstruction work that rarely reaches the newspapers. The crash’s public argument was conducted in the indicative mood, verdict first and evidence as illustration. Its scholarly argument is conducted in the subjunctive, estimate by estimate, and the difference in mood is one reliable marker separating the two literatures when a reader holds them side by side.

A field guide for the next crisis commentary

The competences this case study builds are transferable, because the rhetorical structure of crisis commentary recurs even when the instruments change. When the next severe financial episode arrives, as the frameworks surveyed here would all lead one to expect eventually, the commentary will again manufacture prophets and again announce vindications within weeks. A reader formed by the present comparison has a checklist against that hour. Ask first for the documented sequence, in dates and institutions, before accepting any causal vocabulary. Ask which explanation, of the several the favoured tradition contains, is actually being advanced, and decline the definite article when it is offered without the selection being acknowledged. Ask what the explanation takes the growth of the relevant financial sector to have been, symptom, mechanism, destination or sideshow, since that answer locates it in the comparative field faster than its slogans do. Ask what observation would embarrass it, and treat evasiveness at that point as the answer. Ask whether the warnings being celebrated specified mechanism and horizon or merely orientation, and price them accordingly. Ask, finally, at which level the account operates, proximate mechanism, enabling condition or structural driver, and resist any scoring that grades accounts at different levels as if they had collided when they may have passed, or as if they had harmonised when their claims to primacy collide directly.

Applied retrospectively, the checklist reveals how much of the post 2008 commentary fails on procedure before it fails on substance. The vindication features fail the second and fifth questions. The greed editorials fail the first, offering a motive in place of a sequence. The deregulation slogans fail the third, mistaking a permission for an account of what the permitted growth was. The strongest works in every tradition pass most of the checklist most of the time, which is why the checklist sorts better than allegiance does. There is a civic use in this beyond the seminar: electorates asked to judge rescue packages, regulatory reforms and the officials who design them are being offered, implicitly, an explanation of the crisis those measures answer, and the quality of the public decision cannot exceed the quality of the explanation smuggled inside it. A readership that prices claims of vindication correctly is harder to stampede, in either direction, when the next September arrives.

Why the rivalry matters beyond the tradition

It would be possible to treat the four way Marxist dispute as an internal affair, of interest to the tradition’s members and its historians, and the treatment would be a mistake. The questions the four accounts divide over are the questions any adequate political economy of the crash must answer in some vocabulary: whether the pre crash order’s difficulty lay in production or in distribution, in a surplus of capital or a shortage of demand, in national policy error or in the structure of international competition. Mainstream economics divides over the same terrain under different flags, in the arguments between those who model the episode as a financial shock befalling a fundamentally serviceable system and those who model the financial system as the predictable product of the distribution and the demand regime beneath it. The Marxist rivalry is therefore best read as one theatre of a general unsettlement, distinguished by the age of its frameworks and the explicitness of its mutual criticism. Its value to an outside reader lies precisely in the explicitness: positions that elsewhere circulate as modelling choices or as temperaments appear here as named doctrines with named exponents, forced by their rivalry to state expectations and face counter evidence. A student who learns the general unsettlement through this theatre acquires the map and the legend together, and can afterwards find the same positions in literatures that never mention Marx, wearing the dress of central bank speeches, treasury papers and prize winning monographs.

The great migration of debt: from bank balance sheets to public ones

One of the crash’s most consequential and least theorised features was the relocation of its debts. The obligations began as household mortgages, were transformed into securities held across the financial system, threatened the solvency of banks and insurers, and ended, through rescues, guarantees and recessions, as public debt on the balance sheets of states, where they remained for the following decade as elevated ratios of government borrowing to national income. Each migration changed the debt’s politics. A mortgage default is a private misfortune processed by contract. A bank failure is a systemic emergency processed by officials over a weekend. A public debt burden is a distributional argument processed by parliaments over years, in which the questions become who pays, through which taxes, which services and whose pensions, for losses that originated in transactions most taxpayers never saw. The consolidation era that followed in much of Europe was, in this light, the crash’s third act: the conversion of a financial crisis into a prolonged argument about the size and obligations of the state, conducted in the vocabulary of fiscal rectitude while its balance sheet origins faded from memory.

The four accounts read the migration as predictably as they read everything else, and the readings expose their theories of the state in miniature. The profitability reading treats public borrowing as the system’s shock absorber of last resort, the state warehousing claims that private balance sheets could no longer carry, and postponing the validation question the crash had posed. The underconsumption reading treats the austerity that followed as the distributional settlement reasserting itself, demand suppressed a second time by policy after being suppressed once by the debt collapse, and points to the stagnation of the consolidation economies as the mechanism’s signature. The overaccumulation reading watches the state underwrite the next absorption, infrastructure programmes and property support, and asks where the surplus will be sent this time. The long downturn reading treats the debt argument between nations as the competitive squeeze conducted by fiscal means. The mainstream fiscal literature, for its part, produced careful work showing that the debt ratios were driven more by collapsed revenues and rescue costs than by discretionary stimulus, a finding that acquitted the emergency response and indicted the consolidation timing in several countries, and that sits comfortably beside the underconsumption reading while owing nothing to it. The migration itself, as a documented fact, is neutral; what a reader makes of it depends on the framework carried to it, which is the case study’s recurring lesson wearing fiscal dress.

How the episode is taught, and what the teaching selects

The crash entered syllabuses within two years of the event, in economics, politics, sociology and history courses, and its pedagogical afterlife now shapes what most educated readers believe about it. Teaching selects. A course organised around market failure teaches the crash as the supervisory and incentive accounts present it, with the instruments in the foreground and a reform agenda as the natural conclusion. A course organised around the history of economic thought teaches it as the return of repressed doctrines, Keynes for the policy response, Minsky for the dynamics, Marx for the deep background, with the texts in the foreground. A course within the Marxist tradition teaches it, inevitably, through whichever of the four accounts the instructor holds, and students may leave believing the tradition possesses one explanation because their syllabus contained one. The case for teaching the episode comparatively, as this article has practised, is that the event is unusually well documented, the rivals are unusually articulate, and the evidence discriminates among them unusually often, so that students can be shown explanation itself at work rather than being handed its results. The four accounts and the six mainstream positions between them exhibit every gradation of level, mechanism and depth that the philosophy of social science catalogues, attached to a single narrative that a class can hold in common. Few episodes in the social sciences offer a laboratory as clean, and the condition of the laboratory, a rich public record produced by inquiries conducted for other purposes, is itself worth a lecture: what a society chooses to investigate about its catastrophes determines what its students will be able to think about them.

Standings: what each account keeps, and what it must give up

A comparison that never renders standings has withheld its conclusion, so the standings follow, stated as the evidence presented in this article supports them and no further. The profitability account keeps the deepest question in the field, why capital sought financial returns so persistently across so many jurisdictions, and keeps a serious empirical programme addressed to it. It gives up, or should give up, any claim to explain the crash’s timing and mechanics, and it carries an unresolved measurement dispute at its centre that its exponents have an obligation to narrow rather than to cite selectively. The underconsumption and financialisation account keeps the household balance sheet, the distributional prehistory and the shape of the slow recovery, the three bodies of evidence it fits with least strain. It gives up the corporate sector’s behaviour, which it explains least well, and it owes a cleaner statement of its counterfactual, the stagnation that debt supposedly postponed, in a form evidence could engage. The overaccumulation account keeps the sectoral and geographical pattern, the property form of the boom in country after country, and a theory of why the pattern recurs. It gives up precision about its central magnitude, the surplus, whose independent measurement remains the programme’s unfinished business. In the standings, the long downturn account keeps the international frame, the manufacturing prehistory and the long aftermath, and it keeps the distinction, valuable in itself, between a crisis of a conjuncture and a downturn of an epoch. It gives up the aggregate American profit data in their plainest reading, which it must work harder than its rivals to accommodate, and it shares with the profitability account a timing debt it pays through the credit mechanisms of others.

The mainstream standings can be rendered in the same spirit. Minsky keeps the mechanism, the finest anticipatory description of financial fragility in the literature, and gives up, by his framework’s own design, an account of the background conditions that loaded the mechanism in this instance. The savings glut keeps the level of world interest rates and the international symmetry of the boom, and gives up the domestic manufacturing of toxic assets, which it must import other accounts to explain. The supervisory account keeps the institutional record and the reform agenda, and gives up the generality of the failure, which occurred across too many regimes to be a story about particular referees. The monetary account keeps a genuine contributing condition and a falsifiable form, and gives up sufficiency, since cheap policy money and property booms correlate too imperfectly across countries to bear the whole weight. The incentive account keeps the collapse of underwriting quality, about which it is simply right, and gives up the demand side of the chain. The inequality and credit account keeps the distributional mechanism in a form a finance ministry could operationalise, and gives up, relative to its Marxist cousin, a theory of where the inequality came from.

No line in these standings awards the crash to a tradition. That is the comparison’s finding, and it is a finding rather than a failure of nerve: the episode’s causation is genuinely stratified, the strata are genuinely disputed, and the disputes are genuinely narrower than the rhetorical war conducted over them suggested. What the Marxist tradition contributed, at its best, was a set of disciplined attempts on the deepest stratum, four in number and mutually incompatible, whose rivalry maps the space of possible deep explanations more completely than any one of them maps the truth. What it contributed at its worst was a rhetoric of vindication that concealed the rivalry from its own sympathisers. This article has been written in the conviction that the first contribution is the one worth keeping, and that keeping it requires surrendering the second.

The afterlife of an episode: anniversaries, inquiries and the sediment of interpretation

Public catastrophes acquire sedimentary interpretations, laid down in layers that later readers mistake for the event itself. The crash’s first layer was journalistic and prosecutorial, written in the months of the panic, preoccupied with decisions, personalities and the allocation of blame, drawing on the record available to reporters and on the accounts of participants with reputations to defend. Its second layer was official: the inquiry commissions, the central bank histories, the parliamentary reports, slower, documented, constrained by terms of reference that privileged the supervisory and the proximate. Its third layer was scholarly, the monographs and journal literatures of the succeeding decade, in which the rival explanations examined in this article took their settled forms and the statistical disputes were joined in earnest. A fourth layer is now forming in retrospectives and anniversary assessments, in which the episode is narrated as the opening of a political era, the moment after which trust in institutions, parties and expertise is said to have entered its decline; that layer has its own evidence and its own temptation, which is to read backward from later politics a significance the participants could not have intended and the mechanisms cannot bear.

The sedimentary structure matters for the comparison because each layer privileges the explanations native to it. Journalism rewards the incentive and conduct accounts, whose materials are interviews and documents. Official inquiry rewards the supervisory account, whose materials are the inquirers’ own. Scholarship rewards the deep accounts, whose materials are series and structures, and retrospect rewards whichever account best dramatises the era the retrospective believes followed. A reader who samples only one layer will inherit its privileging invisibly and mistake a property of the archive for a property of the event. The remedy is the discipline practised throughout this article: to ask of every claim which layer’s evidence sustains it, to check the deep claims against the series and the proximate claims against the documents, and to notice when an argument quietly changes layers mid paragraph, borrowing the authority of the inquiry record for a structural conclusion the record was never assembled to test. Interpretations sediment; evidence does not, and the difference between the two is the ground on which honest comparison stands.

Reform after the crash, and the explanations the reforms presupposed

Legislation is frozen explanation, and the reforms enacted after the crash repay reading as arguments about causation wearing statutory dress. The American statute of 2010, the first comprehensive financial reform since the nineteen thirties, created a consumer protection bureau, a council charged with watching systemic risk, a resolution authority for failing financial giants, higher capital and supervision for the largest institutions, a restriction on proprietary trading by deposit taking banks, and a requirement that securitisers retain a share of the risk in what they sold. Each provision answers a diagnosis. Risk retention answers the incentive account. The systemic council and resolution authority answer the perimeter account and the run on shadow banking. The trading restriction answers a conduct story about the corruption of deposit banking by speculation. The consumer bureau answers the underwriting collapse at the borrower’s end of the chain. Read as a whole, the statute is the mainstream mechanism literature enacted: the crash, in its implicit theory, was a failure of incentives, perimeters and supervision inside a system whose deeper configuration required no alteration.

The European measures tell the same story in a different institutional accent: a banking union constructed in stages, with supervision of the largest banks moved to the central bank and a common resolution mechanism behind it, capital rules tightened across the union, and the sovereign bank embrace addressed, as yet incompletely, by treating banks’ holdings of their own governments’ debt as the exposure it had shown itself to be. The international framework raised the quantity and quality of capital the largest banks must hold and added liquidity requirements aimed squarely at the funding fragility of 2008. The Marxist accounts read this entire architecture with a consistent scepticism that follows from their diagnoses: if the pressure to expand credit arises from profitability, distribution or accumulation dynamics that reform leaves untouched, then repaired perimeters relocate fragility rather than abolishing it, and the location of the next episode will be found wherever the new rules bind least. That prediction is testable in principle and uncomfortable in practice, since its confirmation would arrive as another crisis and its falsification requires decades of calm. The mainstream defence is equally consistent: mechanisms that failed have been rebuilt, runs of the 2008 form are structurally harder, and the demand for a reform that also transforms distribution or accumulation asks financial legislation to perform the work of an entire political programme. A reader grading the reforms learns to hold both: the statutes embody real knowledge about the proximate disaster and genuine silence about its background, and the silence is not hypocrisy but the boundary of what the enacting coalitions were assembled to do. Explanation and remedy part company at exactly that boundary, which is why the rivalry among explanations survives the reforms it helped to shape. The statutes also supply a quiet test of the supervisory account itself: that account predicts that a system carrying more capital, wider perimeters and resolution powers should withstand a housing correction of the 2006 magnitude without a general panic, and the prediction is well formed enough that a future episode could genuinely embarrass it, which is more than can be said for several grander claims examined in this article.

Two famous verdicts, graded against the record

The anniversary literature settled into two verdicts so familiar that they now function as the episode’s common sense, and the comparison’s standards can be applied to them as a final exercise. The first verdict holds that the crash discredited the idea of self regulating markets once and for all, and that everything since has been the slow acknowledgement of that discrediting. Graded carefully, the verdict is half earned. The specific doctrine it targets, that financial markets left to themselves price risk correctly and police their own stability, did suffer a documented refutation: the instruments were mispriced by orders of magnitude, the policing institutions sold the mispricing, and the state’s rescue conceded the system’s dependence on public support at the moment of truth. But the verdict overreaches when it extends the refutation to every claim ever made for market coordination, a scope the episode cannot test, and it quietly substitutes a mood of settled disillusion for the unsettled question of which alternative arrangements the record supports. The Marxist accounts divide over exactly that question, which is why the verdict, even in its earned half, selects none of them.

The second verdict holds that nothing changed: that the rescued system resumed its practices, its scale and its rewards essentially intact, and that the crisis therefore stands as a pure demonstration of political power, the financial sector’s capacity to survive its own catastrophe unreformed. Graded against the record, this verdict too divides. In its favour: the largest institutions emerged larger through crisis mergers, the structure of remuneration recovered, financialisation by several measures resumed its growth, and the reforms enacted were negotiated to a mildness their original proponents acknowledged. Against it: capital requirements did rise substantially, a resolution apparatus for failing giants was constructed where none existed, the shadow banking perimeter was drawn wider, proprietary trading was restricted in the largest jurisdiction, and the supervisory state that emerged employs more people with wider powers over more of the system than the one that failed in 2008. Whether those changes amount to transformation or to fortification of the same order is a genuine question on which the four accounts divide along their familiar lines, the mechanism readers seeing a repaired machine, the structural readers seeing a machine repaired precisely so that its deeper dynamics can resume. The verdict of pure stasis is thus as much a selection as the verdict of total discrediting, and a reader who repeats either without the grading has learned the anniversary mood rather than the episode. The case study’s last word is therefore the one its method has carried from the beginning: the crash is large enough to be evidence for many stories and patient enough to discriminate among them, provided the discrimination is actually performed, account by account, stratum by stratum, with the embarrassment test applied to friends first.

For students and exam candidates

This episode is the standard applied case in examinations on Marxist crisis theory, and it rewards a structure rather than a narrative. Open by separating the documented sequence from its explanations: examiners credit candidates who narrate the securitisation chain, the institutional failure and the freeze accurately before theorising any of it. Then present the four Marxist accounts as rivals, one paragraph each, naming the primary cause and the reading of financialisation proper to each. Follow with the prediction question handled at its honest strength, fragility identified, timing unforeseen, comparable warnings outside the tradition acknowledged; answers that claim prophecy lose marks to answers that grade the claim. Close with the strongest objection, the proximate mechanism argument, and the concession it earns. Candidates who reproduce a single account as the Marxist view, however fluently, cap their own ceiling, because the question the case really sets is comparative.

For teachers and seminar leaders

The comparison table is built for classroom use: distribute it before revealing which rows are Marxist, and let students discover that the mechanism rows and the cause rows sort differently than they expected. A productive seminar exercise assigns each participant one account and one shared fact from the worked comparison, requiring the participant to argue the fact as central and then as derivative from a rival’s position; the exercise teaches levels of explanation faster than any lecture on the topic. Set the prediction question as a source exercise, primary warnings from inside and outside the tradition read side by side, and let the class formulate the honest verdict themselves before offering the article’s. Guard the two failure modes of discussion in this area: the session that becomes a tribunal on the banking industry, and the session that becomes a celebration of the tradition. The case is rich enough to punish both.

For researchers, historians, and journalists

For researchers, the live frontier is measurement: the profitability dispute remains open because conventions differ, and work that states its capital valuation, depreciation and sector boundary choices explicitly, and reports results under rivals’ conventions as well as its own, moves the debate more than another restatement of position. Household debt microdata, underexploited relative to the aggregate series, offers the underconsumption account its sharpest possible tests, particularly on the distribution of borrowing across the income scale. Historians should note how much of the evidentiary base was created by the official inquiries and how their terms of reference have quietly bounded the questions the literature asks; the archives of the rescued institutions, as they open, will revise the proximate narrative first and the deep one only later. Journalists covering anniversaries or the next episode entirely can borrow the checklist in this article wholesale: sequence before vocabulary, the named account before the tradition, the embarrassment test before the vindication quote. Keeping structured notes on which account each source represents, and on what evidence would move each camp, is exactly the sort of record that the study notes companion at VaultBook Marxism study notes is designed to hold, and an hour spent building the grid before the interviews will repay itself in every conversation that follows.

Frequently Asked Questions

Q: What caused the global financial crash?

At the level of mechanism, the documented chain is settled in outline: a decade of loosening mortgage credit in the United States generated loans dependent on rising house prices; securitisation pooled and tranched those loans into securities whose ratings rested on models that treated a national price fall as nearly impossible; leverage and overnight funding spread the securities through banks, insurers and money funds worldwide; and when prices fell and defaults rose, the securities became unpriceable, funding evaporated, a major investment bank failed in September 2008, and credit to the wider economy froze. What remains disputed, and what this article compares, is the level beneath the mechanism: why credit had to expand so fast, why housing carried it, why wages, profits and savings were configured as they were. Rival explanations, Marxist and mainstream, are claims about that deeper level.

Q: How do Marxists explain the financial crash?

They explain it in at least four incompatible ways, and the plurality is the point a careful answer has to start from. The profitability account holds that weak returns in production over decades drove capital into financial channels, so the crash surfaced a long accumulation problem. The financialisation and underconsumption account holds that stagnant wages suppressed demand, household debt substituted for income, and the substitution collapsed with house prices. The overaccumulation account holds that surplus capital was absorbed into property and the absorption mechanism broke. The long downturn account locates the cause in international manufacturing overcapacity built up through competition since the late nineteen sixties. Each names a different primary cause and reads the growth of finance differently, as symptom, bridge, destination or sideshow, so no single paragraph can state the Marxist explanation without choosing among rivals.

Q: Did Marxists predict the crash?

No, if prediction means foresight of the event, its mechanism and its timing stated in advance in a form that could have failed. Some Marxist economists had warned for years that the financial expansion was fragile, identifying the dependence of demand on household debt and asset appreciation, and one influential analysis published a decade earlier argued that the world economy was stuck in a long downturn bound to generate instability. Those warnings were genuine and they oriented their readers toward the balance sheets that mattered. But the timing was foreseen by nobody, and comparably specific warnings came from outside the tradition: a housing economist’s documented bubble analysis, economists at the Bank for International Settlements arguing that credit booms were building systemic risk, and a 2005 conference paper on incentives and tail risk delivered to the world’s central bankers. The defensible claim concerns frameworks, not prophecy.

Q: Was the crash caused by falling profitability?

That depends on accepting one of the four Marxist accounts and its measurement conventions, so the honest answer is that the claim is contested inside the tradition as well as outside it. Writers in the profitability family present long statistical series in which the rate of profit in the advanced economies, and especially in the United States, never durably regained its postwar levels, with investment chronically weak relative to the capital stock and debt filling the gap. Critics answer that profit shares and several reconstructed rate series did recover substantially from their trough, that the variable moved far too slowly to detonate in a particular autumn, and that the acute phase of any profitability story has to be borrowed from the credit dynamics other accounts place at the centre. The dispute turns on capital valuation and sector boundaries, and it remains open.

Q: Was the crash caused by inequality and household debt?

Inequality and household debt form the strongest demand side account of the crash, held in variant forms by one Marxist family and by a prominent mainstream economist alike. The mechanism is documented: typical wages stagnated for decades while top incomes and the costs of housing, health and education rose; borrowing bridged the gap, encouraged by policy that preferred credit expansion to redistribution; household debt reached unprecedented ratios to income; and the bridge depended on house prices, so its collapse transmitted directly into consumption and into the mortgage securities. The cross country pattern supports the account, since the deepest banking crises struck where household debt and housing booms had grown fastest. Its exposed flank is the corporate sector, where investment weakness coexisted with high profits and large cash balances, a combination the account explains less comfortably than its household evidence.

Q: What role did financialisation play in the crash?

Financialisation names the growth of finance relative to the rest of the economy, and the crash literature divides over whether that growth was cause, mechanism or symptom. In the profitability account it is a symptom, capital’s refuge from disappointing returns in production. In the underconsumption account it is a mechanism, the machinery by which debt replaced wages in sustaining demand. In the overaccumulation account it is the plumbing that delivered surplus capital to the built environment. Minsky’s framework treats it as an endogenous product of stability itself, risk postures migrating toward fragility during calm. What no serious account permits is using the word as a complete explanation, since the expansion of finance is precisely what requires explaining. The question to put to any writer is which of those roles the term is playing in the sentence where it appears.

Q: Do Marxist explanations of the crash conflict?

Yes, structurally and not merely in emphasis. The four accounts identify different primary causes, weak profitability, suppressed demand, surplus absorption into property, and international overcapacity, and a primary cause is a claim about what the event cannot be explained without. They read financialisation in four different roles, predict different obstacles to recovery, and prescribe divergent remedies, from distributional repair to scepticism that any remedy within the existing structure reaches the difficulty. They also dispute one another’s evidence directly, most sharply over whether profitability recovered before the crash. What they share is a negative thesis, that the collapse was no accident befalling a sound order, and a preference for slow variables over sudden vices. Presenting them as one explanation with four chapters misrepresents the literature and conceals the choices a reader actually has to make among research programmes.

Q: What is Minsky’s explanation and is it compatible?

Hyman Minsky’s financial instability hypothesis holds that stability breeds fragility from within: during calm periods borrowers and lenders accept thinner margins of safety, financing migrates from hedge postures, where cash flow covers all obligations, through speculative postures dependent on refinancing, toward Ponzi postures dependent on rising asset prices, until a disappointment triggers forced selling and collapse. The hypothesis describes the migration of the two thousands with uncomfortable fidelity, including the role of innovation in evading margins of safety. Minsky was a heterodox Keynesian, not a Marxist, and his framework contains no theory of value or class. Compatibility is therefore level specific: his dynamics can be, and often are, accepted by Marxist writers as the mechanism through which deeper conditions in profitability, distribution or accumulation expressed themselves, while the foundations remain distinct. His account also blocks any claim that detailed anticipation of the mechanism was a Marxist monopoly.

Q: What was the long downturn thesis?

The thesis, associated above all with Robert Brenner and stated in a long analysis published in 1998, holds that the advanced economies entered a sustained downturn from the late nineteen sixties because world manufacturing developed chronic overcapacity. Successive waves of industrial entry, first from rebuilt Germany and Japan, then from East Asia and eventually China, expanded supply faster than remunerative demand, while established producers carrying heavy sunk costs declined to exit, so prices, margins and profitability were squeezed across the system for decades. Governments and households responded with borrowing that sustained demand artificially, and the asset booms of the nineteen nineties and two thousands were, on this reading, props within the downturn rather than signs of health. The crash was the props failing. The thesis is admired for its international frame and contested, including within Marxism, over its reading of profitability data for the years after the mid nineteen nineties.

Q: Did the crash vindicate Marx?

Not in any form that survives the comparison this article performs. Marx wrote penetratingly about credit, fictitious capital and the crisis as a forced reconciliation of claims with the means of payment, and those materials illuminate 2008 genuinely. But vindication of a thinker by an event requires the event to confirm a distinctive claim against its rivals, and the aftermath produced four incompatible Marxist explanations alongside strong non-Marxist ones, several of which describe the mechanism better. An event consistent with all of them vindicates none of them in particular, which is the vindication trap. What can fairly be said is that the tradition contained frameworks that took financial fragility seriously while triumphalist commentary dismissed it, that its questions about profitability and distribution remain live in the scholarly literature, and that its standing after the crash is a matter of research programmes graded against evidence, not of a verdict delivered by a single autumn.

Q: How does the explanation comparison separate Marxist accounts of the crash?

The comparison separates them by forcing every account, Marxist and mainstream alike, to answer the same five questions in a form that exposes disagreement. What is the primary cause alleged? What was the expansion of finance, symptom, mechanism, destination or byproduct? What follows for the character of the recovery? What evidence would count against the account? And at which level does it operate, proximate mechanism, enabling condition or structural driver? Marxist accounts that sound identical at the level of slogan, all of them critical of finance and of the pre crash order, separate cleanly under the questions: their causes are different variables, their financialisations have different roles, their expected recoveries diverge, and their embarrassment conditions point at different data series. The separation is the table’s purpose, and it generalises to any crisis literature a reader meets afterward.

Q: What is the vindication trap in claims that Marx predicted the crash?

The trap is the inference from an event’s consistency with a tradition to the tradition’s confirmation, performed while ignoring that the tradition contains rival accounts which cannot all be true. A crash is consistent with a falling profit rate story, a wage stagnation story, an overaccumulation story and an overcapacity story at the loose level of narrative, because each was constructed partly with such events in view. Consistency therefore discriminates among none of them, and the rhetoric that treats the episode as a coronation selects one account silently, usually the writer’s own, and presents it as the whole tradition’s verdict. The trap has a hostile mirror image in which the rivalry itself is treated as proof of incoherence. Escaping both requires the slower work of stating each account’s expectations and grading them against the record, which converts applause and dismissal alike into comparison.

Q: Why did some accounts centre subprime lending and others the profit rate?

Because the two starting points encode different theories of where explanation lives. Accounts centred on subprime lending begin with conduct and instruments: the deterioration of underwriting, the fee chain that rewarded it, the models that mispriced its risk. They explain the quality collapse of the loans and the sequence of the panic better than any rival, and they leave open why millions of such loans became profitable to make in the same half decade. Accounts centred on the profit rate begin with structure: the long behaviour of returns on capital, which they treat as the pressure that drove funds toward finance and households toward debt. They address the timing of the era rather than of the autumn, and they leave the instruments as plumbing. Narrative choice follows explanatory commitment, and the strongest works in the literature are those that begin in one scene and earn their way to the other without pretending the journey was unnecessary.

Q: How should a historian weigh a crisis narrative written during the crash?

With respect for its texture and suspicion of its facts. Narratives written in 2008 and 2009 preserve what the uncertainty felt like from inside, the genuine possibility contemporaries assigned to outcomes that did not occur, which hindsight erases; that makes them irreplaceable evidence about perception and decision. The same narratives guessed magnitudes under pressure, allocated blame before records were available, and fitted causes to statistics later revised beyond recognition. The working rules are to separate strata, contemporaneous narrative for perception, official inquiries for documented sequence, statistical reconstruction for the slow variables; to privilege each source for what it was positioned to know; and to read silences as evidence, since what a narrative omits, wages in one, the repo market in another, maps its framework’s blind spots. A narrative’s date of composition is part of its evidentiary identity, neither a disqualification nor a credential.

Q: What did the crash change in academic Marxist economics?

Less than the publicity announced and more than its critics concede. The measurable changes were threefold. Finance moved from a derivative topic toward the centre of the field, with work on financialisation, household financial expropriation and shareholder value becoming standard reference points. The empirical disputes sharpened, because the obligation to explain a definite event forced the profitability camps into more explicit measurement conventions than polemic had previously required. And a generational shift followed, with scholars formed by the episode treating crisis as a recurrent object within a continuing order rather than as a terminal harbinger, which redirected energy toward instability, distribution and state management. What did not happen is equally definite: the four rival accounts did not converge, the discipline of economics did not change its foundations, and the tradition’s centre of gravity settled into research programmes rather than into the vindicated unity of newspaper imagination.