Dalton Rongxuan Zhang

The conventional view of the aggregate effects of market power focuses on product markets. We develop a complementary view, where the distributional effects of market power result in novel macroeconomic consequences because they alter supply and demand in asset markets and the long-run equilibrium cost of capital. We develop a tractable model framework with heterogeneous households and identify three forces through which market power affects asset markets: market power (i) reduces firms’ demand for productive capital, but also (ii) redistributes income from low-saving workers to high-saving firm owners, and (iii) raises the market value of equity claims on public firms. Together, these forces determine the endogenous long-run response of interest rates, capital accumulation, and output to changes in market power. Using U.S. data, we find that rising markups from 1980 to 2018 lowered interest rates by 1 pp, reduced the labor share by 5.6 pp, and increased the top 1% wealth share by 12 pp in line with empirical estimates. The resulting decline in the cost of capital mitigates half of the 11% long-run output loss implied by a representative-household benchmark. Although output losses are smaller, aggregate utilitarian welfare loss is five times larger, reflecting greater inequality and losses borne by working households with higher marginal utility of consumption. We extend our framework and consider other sources of market power, such as labor market power, and multiple asset classes, relating the rise of market power to a larger equity premium and the decline of risk-free rates.

The possibility of fiscal dominance in the representative-agent New Keynesian model (RANK) hinges on the assumption that income is perpetually demand-determined: fiscal deficits can drive output and inflation within that model only insofar as they trigger infinitely lasting, self-sustained shifts in aggregate spending and income. Moving to heterogeneous-agent New Keynesian models (HANK) opens the door to a different pathway: classical non-Ricardian effects, due to finite horizons or liquidity constraints. A refinement motivated by the model’s intended focus on short-run phenomena—requiring a return to flexible-price outcomes in finite time—arrests the infinite feedback loop between spending and income, leaving only the classical non-Ricardian mechanism, and makes sure that the study of monetary-fiscal interactions is not centered on hard-to-test assumptions regarding beliefs at infinity.

We develop a structural approach to identify latent components in households’ macroeconomic beliefs. We use the covariance structure of cross-sectional disagreement across macroeconomic expectations, combining sign and narrative restrictions based on mass-shooting exposure. This identifies pessimism, demand, and supply components and provides a unified anatomy of expectation comovement while remaining agnostic about whether it reflects incomplete information or subjective views. Pessimism, a broad adverse shift in beliefs, accounts for a large share of covariance alongside supply-side views. It is concentrated among individuals with deteriorating finances, volatile income, and poor health, who anticipate greater earnings uncertainty and higher job-loss and default risk. The supply component is more common among older individuals with greater lifetime inflation experience; the demand component is more common among higher-income, more-educated individuals. Overall, the results suggest that studying expectation comovements, moving beyond isolated forecast errors, is important for theories of expectation formation.

High-inflation episodes often feature confusion about the source of aggregate shocks. We develop a theory of this confusion. Agents observe noisy signals about output and inflation and may confuse demand and supply shocks. When inflation is more informative, agents tend to interpret high inflation as signal of both adverse supply and strong demand. This confusion amplifies the inflation response to true supply shocks and dampens the expected-output response to true demand shocks. Estimating the model before and after 1990 reveals a shift from an inflation-signal-dominant to an output-signal-dominant regime, helping explain stronger pre-1990 inflation responses to supply shocks, the flattening of the Phillips curve, and the supply-side view in expectations.

We offer guidance and tools for augmenting macroeconomic models with a realistic friction in expectations and for quantifying its implications. We start by distilling the common essence of a few popular alternatives to FIRE (Full Information Rational Expectations), including noisy or sticky information, rational or behavioral inattention, level-k thinking and cognitive discounting. We then develop a new specification, which captures the same essence while maximizing tractability. We show how this works in both RANK (the Representative-Agent New Keynesian model) and HANK (Heterogeneous-Agent New Keynesian models). We further clarify the separate partial- and general-equilibrium effects of informational frictions or bounded rationality; emphasize the interaction of such frictions with the Intertemporal Keynesian Cross and other forms of strategic complementarity; review concrete lessons for business cycles and macroeconomic policy; and discuss how to discipline the theory with appropriate survey evidence.

This paper revisits conventional wisdom on market power by decomposing it into product and labor components using U.S. Census data. We document: (i) the rise in aggregate market power is driven by wage markdowns, while price markups remain stable; (ii) larger firms have lower product but higher labor market power; and (iii) incomplete price pass-through reflects monopsony power rather than demand curvature. A general equilibrium model with nonparametrically identified demand and supply shows about 30% of rising markdowns stem from productivity dispersion, with the rest from demand-supply shifts and their interaction.

We document three correlated trends among U.S. public firms: longer initial loss spells, rising sales concentration, and a shift from physical investment toward R&D and sales and marketing. Using Compustat and sales and marketing data, we estimate revenue production functions with physical, knowledge, and customer capital and find rising revenue elasticities of knowledge and customer capital. We develop a heterogeneous-firm model in which knowledge capital is non-rival across a firm's customers. Customer investment expands demand and the scale over which knowledge is used. Greater knowledge intensity raises returns to knowledge and customer investment, allowing productive firms to grow for longer before becoming profitable. In a calibrated comparison of stationary distributions disciplined by the estimated elasticity changes, the model raises the sales share of the top 1 percent of firms by 5.5 percentage points and endogenous total factor productivity by 5.1 percent. About 90 percent of the productivity gain comes from the non-rival use of knowledge across customers.

Recent asset pricing research highlights subjective expectations of long-run profit growth as a key driver of return variation. We show that these expectations—and their deviations from full-information rational expectations—systematically depend on firms’ market power. Using U.S. Census LBD data and analyst forecasts, we find that product market power amplifies and prolongs overreaction, while labor market power dampens it. We also document that firm-level overreactions are an order of magnitude larger than aggregate overreactions. Embedding these heterogeneities in a macro-finance model reproduces core asset pricing moments—high excess returns, volatility, and predictable reversals—and shows that distorted expectations, interacting with heterogeneous market power, generate procyclical misallocation and amplify macro-financial fluctuations.