Acemoglu's alternative is that politics was not merely an intervening channel responding to inequality, but an autonomous driving force: lobbying, campaign contributions, and privileged access by the well-off and well-organized shaped the financial industry over 25 years, and those distortions then caused both the financial crisis and much of the top-end inequality observed in the US.

The rise of top earners as wage (rather than capital) earners cannot be explained solely by increased demand for skills; organizational changes in how industries are structured and how top management and finance professionals are incentivized (options, bonuses, and broader changes) must also be responsible.

Rajan's thesis holds that an exogenous, technology-driven rise in inequality created political pressure that led the government to encourage GSEs like Fannie and Freddie to provide cheap, subsidized credit to lower- and middle-income households, fueling a housing bubble whose collapse caused the financial crisis.

There is a crucial distinction between 'labor-market inequality' (gaps measured up to the 90th percentile, e.g., college premium, 90/10 ratio) and 'top inequality' (how much the top 1% or 0.1% earn relative to the rest); the causes and data sources for each differ, and standard survey/census data are top-coded so they cannot reveal what happens at the very top.

Acemoglu remains an optimist about US institutions, noting they are among the strongest in the world, that overt corruption and self-enrichment by politicians have been limited, and that the political system is not stuck or in the pocket of any single group—though the revolving-door lobbying culture is distasteful.

Per Piketty and Saez IRS data, the top 1% of US earners captured about 10% of national income in the 1970s, rising to nearly 25% by the late 2000s—and remarkably, these mega-earners shifted from being rentiers living on inherited capital to being W-2 wage earners.

The consensus in labor economics is that rising labor-market inequality over the last 30 years is led mainly by technology (information technology, robotics, new organizational forms increasing demand for skilled/abstract labor) combined with a slower increase in the supply of skills, with trade playing a role intermediated by technology (offshoring/outsourcing) and institutional factors like union decline and the minimum wage still contested.

Labor-market inequality, while higher now than at any point since World War II, is not out of line with historical trends (e.g., early 20th century US) and is far below levels seen in Latin America; in contrast, the rise in top inequality is historically remarkable.

Research by Thomas Philippon and Ariell Reshef shows that beginning in the 1970s there was a remarkable explosion in wages in the financial sector relative to the rest of the economy, indicating finance plays a disproportionate role in the top 1%.

Allocating capital is genuinely socially valuable, but it is unclear whether profiting from short-run price fluctuations contributes to getting the long-run prices right; if you can make millions from a few hours of mispricing that doesn't materially affect the long-run allocation of capital, that activity's social value is questionable.

The explosion in political action committees, campaign financing, and lobbying expenditures over the last 30-35 years provides a mechanism for why politicians track high-income constituents: influence comes not only from buying politicians but from access—being able to talk to and shape their views—which itself requires money.

Research by Neil Bhutta finds that the Community Reinvestment Act and the GSE underserved-areas goal provisions (setting Fannie/Freddie targets for mortgages in lower-income and minority areas) had only very small effects, suggesting these specific policies were not decisive causes of the housing boom.

The timing of inequality undermines Rajan's hypothesis: the big increases in both 90/10 and 50/10 inequality occurred in the 1980s, while the late 1990s and 2000s were remarkable for the bottom and middle doing reasonably well—so Rajan's story requires an awkward ~20-year delay between the 1980s inequality rise and the crisis.

Even a milder version where bankers get a large cut of profits but bear no downside beyond losing their job (still keeping a respectable salary) creates terrible incentives, because they can make humongous amounts by succeeding at risky bets while not bearing the losses.

Political science evidence (Larry Bartels and others) shows that senators' votes—on both economic and social issues—are highly correlated with the opinions of their high-income constituents (top third), moderately correlated with middle-income constituents, and not at all correlated with low-income constituents, making it implausible that politicians were responding to the bottom's discontent.

There is causal evidence that access to different information sources affects political outcomes: DellaVigna and Kaplan's Fox News study (exploiting staggered cable rollout) finds a sizeable effect on Republican presidential vote share, and other studies (Stromberg-Snyder, Ferguson) show politicians behave differently and lobbies are countered more where media coverage is greater.

Reform should avoid big top-down interventions and instead make existing US political processes work better—curbing the worsening of money in politics, strengthening media's checking role, enacting campaign finance reform, and ensuring broad, informed participation; the Citizens United decision moved in the wrong direction by undermining checks on campaign contributions.

Media is a key check on money-driven politics, but a vicious circle exists: democracy corrects misbehaving politicians via voting, yet campaign contributions enable politicians to influence votes (through advertising and name recognition), so money can blunt the very mechanism meant to discipline it.

Wall Street's influence stems not just from money but from a near-monopoly on financial expertise—possibly greater than its true expertise justifies—so when policymakers like Paulson consult Blankfein or Dimon during critical periods, they receive not objective truth but truth filtered through and shaped to fit those firms' interests.

Sustained economic growth requires a particular political environment: stability, enforcement of property rights, rule of law, a society able to innovate and channel talent into productive activities, and the capacity to reallocate resources from declining low-productivity firms to high-growth ones—the essence of Schumpeterian creative destruction.