Macroeconomics Foundations

Macroeconomics studies aggregate economic outcomes — output, employment, inflation, interest rates, exchange rates, growth, and crises — at the level of nations and the world economy. It synthesizes accounting identities (national income), short-run demand models (IS-LM, AD-AS, New Keynesian), long-run growth theory (Solow, Romer), monetary and fiscal policy frameworks, and the empirical analysis of business cycles and financial crises. This reference covers the canonical models, their originators, current (2024–26) data infrastructure, and the live policy debates of the post-COVID, post-Ukraine, AI-productivity era.

Cross-reference microeconomics-foundations for the household and firm optimization that underpins New Keynesian micro-foundations, corporate-finance-and-markets for the asset-pricing and term-structure side, electricity-markets for the energy supply shocks that drove the 2021–23 inflation cycle, _index for the climate-transition macro literature (green industrial policy, carbon pricing, stranded assets), and hypothesis-testing-mle for the econometric machinery (MLE, Bayesian estimation of DSGE models).

1. National Income Accounting

The three approaches to GDP

Gross Domestic Product (GDP) is the market value of all final goods and services produced within a country’s borders in a given period. It can be measured three equivalent ways — the expenditure, income, and production approaches — which yield the same total (up to a statistical discrepancy) because every dollar spent on a final good is a dollar of income to someone and a dollar of value added in production.

Expenditure approach (the textbook identity). Y = C + I + G + (X − M), where C is private consumption, I is gross private domestic investment (fixed investment plus inventory change), G is government consumption and gross investment, and (X − M) is net exports. In the U.S. national income and product accounts (NIPA) maintained by the Bureau of Economic Analysis (BEA), C is roughly 68% of GDP, I 18%, G 17%, and net exports −3%. Investment is the most volatile component; consumption the smoothest.

Income approach. GDP = compensation of employees + gross operating surplus + gross mixed income + (taxes − subsidies on production and imports). In practice this is computed as national income (labor and capital income) plus consumption of fixed capital (depreciation) plus indirect taxes net of subsidies. The labor share of national income in advanced economies has trended down from ~65% in 1980 to ~58% by 2024 — a stylized fact documented by Karabarbounis and Neiman (2014) and central to the inequality debates of the past decade.

Production (value-added) approach. GDP = sum across industries of gross output minus intermediate consumption. The BEA publishes industry-level value added in the GDP-by-industry accounts; this is the natural starting point for productivity analysis and for the input-output (I-O) tables originating with Leontief (1936, Nobel 1973).

Gross National Income (GNI) = GDP + net primary income from abroad. For most large economies GDP ≈ GNI, but for Ireland the gap is enormous (~25% of GDP) because of multinational profit booking — this is why Irish statistics also publish modified GNI (GNI*) excluding redomiciled firm profits.
Net National Product (NNP) = GNP − depreciation. Conceptually preferable to GDP for welfare analysis but rarely used because depreciation is hard to measure.
Disposable income = GNI − net taxes. Drives consumption in the Keynesian cross.
Nominal vs real GDP: nominal uses current prices; real uses constant base-year prices (BEA currently uses chain-weighted indexes with 2017 as the reference year). The implicit GDP deflator = nominal/real.

Limitations and extensions

GDP omits non-market production (household labor), depletion of natural capital, and is silent on distribution. Stiglitz, Sen, and Fitoussi (2009) led the post-financial-crisis call for “beyond GDP” measures; subsequent work includes the OECD Better Life Index, the UN System of Environmental-Economic Accounting (SEEA), and inclusive wealth accounts. The 2025 G20 finance track has formally adopted SEEA satellite accounts as a reporting requirement for member states.

2. Business Cycles — Stylized Facts

Burns and Mitchell (1946) at the National Bureau of Economic Research (NBER) defined business cycles as recurrent but non-periodic fluctuations in aggregate activity. The NBER Business Cycle Dating Committee (chaired through 2025 by Robert Hall of Stanford) dates U.S. recessions. The two most recent: February 2020 – April 2020 (the briefest on record, two months) and the still-debated mild contraction of 2022 H1 (two quarters of negative real GDP growth but with strong employment, so the committee declined to call it a recession).

Stylized facts (Stock and Watson 1999, updated through 2025):

Consumption is procyclical and less volatile than GDP (smoothing, per the permanent income hypothesis of Friedman 1957).
Investment is procyclical and 3–4× more volatile than GDP.
Employment is procyclical and lagging; unemployment is countercyclical.
Real wages are weakly procyclical; nominal wages are sticky downward.
Inflation is procyclical at short horizons but the relationship has weakened post-1990 (the flattening of the Phillips curve).
The yield curve (10y − 3m or 10y − 2y) inverts before recessions — a stylized fact going back to Estrella and Mishkin (1996). The 2022–23 inversion was the deepest since 1981 and preceded a soft landing rather than a recession, prompting a literature on why the signal weakened.

3. Aggregate Supply and Aggregate Demand (AS-AD)

The AS-AD framework, developed in textbook form by Mankiw and Romer in the 1980s, plots the price level P against real output Y.

Aggregate demand (AD) slopes down: higher P reduces real money balances (Pigou/Keynes wealth effect), raises real interest rates (Keynes effect), and appreciates the real exchange rate (Mundell-Fleming). AD shifts with fiscal policy (G, T), monetary policy (M, i), expectations, and the global cycle.

Short-run aggregate supply (SRAS) slopes up: with sticky nominal wages or prices, higher P raises producer real margins and induces more output.

Long-run aggregate supply (LRAS) is vertical at potential output Y*, determined by the labor force, the capital stock, and total factor productivity (TFP). LRAS shifts with technology, demographics, immigration, and capital deepening.

Equilibrium: SRAS ∩ AD gives the short-run price level and output; convergence to LRAS happens as wages and prices adjust. A negative supply shock (oil, 1973; 1979; 2022 Russia-Ukraine) shifts SRAS left, raising P and lowering Y — stagflation. A demand shock (2020 COVID lockdowns, 2009 deleveraging) shifts AD left, lowering both P and Y.

The 2021–22 inflation surge is best read as a sequence of overlapping shocks: positive AD (fiscal stimulus, suppressed services consumption shifting to goods), negative SRAS (supply chain bottlenecks, semiconductor shortage, energy shock), and a temporary steepening of the Phillips curve as labor markets tightened.

4. IS-LM and Mundell-Fleming

IS-LM

Hicks (1937, “Mr. Keynes and the Classics”, Nobel 1972) formalized Keynes’s General Theory as the intersection of two curves in (Y, i) space:

IS curve: goods-market equilibrium. Y = C(Y − T) + I(i) + G + NX. Investment falls in i, so higher i lowers Y. IS shifts right with G ↑, T ↓, or autonomous demand ↑.
LM curve: money-market equilibrium. M/P = L(i, Y). Higher Y raises money demand; for given M/P, equilibrium i must rise. LM shifts right with M/P ↑.

IS-LM survives in modern textbooks as a pedagogical tool but has been superseded for research by New Keynesian DSGE models (Section 6). Romer (2000) proposed the IS-MP (Monetary Policy) replacement that treats the central bank’s interest-rate rule as primary, sidestepping money demand — this is now standard in undergraduate texts (Mankiw, Jones).

Mundell-Fleming and the impossible trinity

Mundell (1963, Nobel 1999) and Fleming (1962) extended IS-LM to an open economy with capital flows. The BP curve (or FX curve) represents balance-of-payments equilibrium given the exchange-rate regime.

Impossible trinity (Mundell trilemma): a country can pick only two of (i) free capital mobility, (ii) a fixed exchange rate, (iii) independent monetary policy. The U.S. has (i) and (iii); Hong Kong’s currency board has (i) and (ii); China has (ii) and (iii) (partly, via capital controls). The euro area collectively forfeited (iii) at the country level when joining the single currency — the unresolved fiscal-versus-monetary tension exposed by the 2010–12 sovereign debt crisis.

Under flexible exchange rates and high capital mobility, fiscal policy is weak (crowding out via currency appreciation) and monetary policy is potent. Under fixed exchange rates, the reverse: monetary policy is constrained to defend the peg, and fiscal policy is potent. This Mundell-Fleming logic is central to the IMF’s policy advice to emerging markets.

5. The Phillips Curve

Phillips (1958) documented an empirical negative relationship between U.K. unemployment and nominal wage growth (1861–1957). Samuelson and Solow (1960) translated this to U.S. inflation-unemployment and suggested it as a policy menu.

Friedman (1968, Nobel 1976) and Phelps (1967, 1968, Nobel 2006) demolished the “menu” interpretation by introducing expectations. The expectations-augmented Phillips curve:

π = π^e − β(u − u*) + ε

where π^e is expected inflation and uis the natural rate (later renamed NAIRU, the non-accelerating inflation rate of unemployment). In the long run, π^e = π and u = u, so the long-run Phillips curve is vertical. Monetary policy cannot exploit a permanent inflation-unemployment trade-off — only a transitory one along surprise inflation.

The 1970s stagflation confirmed Friedman-Phelps. By the 1980s, the rational-expectations critique (Lucas 1976, Nobel 1995; Sargent, Nobel 2011) tightened the argument: systematic monetary policy is foreseen, and only the unanticipated component moves real activity.

Modern New Keynesian Phillips curve (NKPC), derived from Calvo (1983) staggered price-setting:

π_t = β E_t π_{t+1} + κ x_t

where x_t is the output gap and κ depends on the frequency of price adjustment. This forward-looking version has been the workhorse of policy DSGE models (Smets-Wouters 2007).

The Phillips curve flattened dramatically from ~1990 through 2019. Hooper, Mishkin, and Sufi (2020) document the slope dropping from −0.6 in the 1970s to near zero by the 2010s — attributed to anchored expectations, globalization, and labor market changes. The 2021–23 inflation surge re-steepened it temporarily, reviving the empirical literature: Bernanke and Blanchard (2023) decompose the surge into supply shocks (dominant), wage-price spirals (limited), and expectations (well-anchored throughout, validating modern central bank credibility).

6. New Keynesian DSGE and the Taylor Rule

The three-equation NK model

The canonical New Keynesian model (Clarida, Galí, and Gertler 1999; Woodford 2003 Interest and Prices; Galí 2008 Monetary Policy, Inflation, and the Business Cycle) reduces to:

Dynamic IS: x_t = E_t x_{t+1} − (1/σ)(i_t − E_t π_{t+1} − r*_t), derived from the household Euler equation with intertemporal elasticity 1/σ.
NK Phillips curve: π_t = β E_t π_{t+1} + κ x_t, from Calvo pricing.
Monetary policy rule (Taylor 1993): i_t = r*+ π_t + ϕ_π (π_t − π*) + ϕ_x x_t, with ϕ_π > 1 (the Taylor principle) for determinacy.

Smets and Wouters (2003, 2007) estimated a richer medium-scale DSGE on euro area and U.S. data with habits in consumption, investment adjustment costs, wage stickiness, and seven structural shocks. Variants are used at the Fed (FRB/US, EDO), ECB (NAWM), Bank of England (COMPASS), and the IMF (GIMF, GFM).

Critiques of DSGE:

Linearization around steady state misses nonlinearities (ZLB, financial crises). Solution methods (perturbation, projection, deep learning) are an active area.
Representative agent obscures distribution. HANK (heterogeneous-agent New Keynesian) models — Kaplan, Moll, and Violante (2018) — restore micro heterogeneity and have changed how we think about monetary policy transmission, fiscal multipliers, and inequality.
Financial frictions absent from early DSGE were a casualty of 2008. Bernanke, Gertler, and Gilchrist (1999) financial accelerator; Gertler and Karadi (2011) banks; Brunnermeier and Sannikov (2014) macro-finance — all are now in the standard toolkit.

The Taylor rule in practice

Taylor’s (1993) original calibration: i = 2 + π + 0.5(π − 2) + 0.5x. Critics argue the Fed has often deviated, especially during the 2003–05 “Greenspan put” period (rates too low, fueling the housing bubble — Taylor’s own retrospective critique) and during 2021–22 (rates lagging inflation by 4+ percentage points, the largest deviation since the late 1970s). The Fed’s official framework since 2020 is “flexible average inflation targeting” (FAIT), reviewed in 2025 — the review concluded FAIT had contributed to the 2021–22 inflation overshoot and signaled a return to a more symmetric inflation-targeting stance for the 2026–30 framework.

7. Monetary Policy and Central Banking

Major central banks and their mandates

Central bank	Mandate	Target	Policy committee	Chair/Governor (2026)
U.S. Federal Reserve	Dual: maximum employment + price stability	2% PCE inflation	FOMC (19 members, 12 voting)	Jerome Powell (term ends May 2026); Lael Brainard succession announced March 2026
European Central Bank (ECB)	Primary: price stability (2% HICP, symmetric)	2% HICP	Governing Council (26 members)	Christine Lagarde
Bank of England (BoE)	Price stability + supporting employment & growth	2% CPI	MPC (9 members)	Andrew Bailey
Bank of Japan (BoJ)	Price stability	2% CPI	Policy Board (9 members)	Kazuo Ueda
People’s Bank of China (PBoC)	Multiple: stability, growth, employment, BoP	Implicit ~3% CPI	Monetary Policy Committee (advisory)	Pan Gongsheng
Reserve Bank of India (RBI)	Flexible inflation targeting	4% CPI (±2%)	MPC (6 members)	Sanjay Malhotra
Swiss National Bank (SNB)	Price stability	<2% CPI	Governing Board (3 members)	Martin Schlegel
Bank of Canada (BoC)	Inflation target	2% CPI (1–3% range)	Governing Council	Tiff Macklem

Conventional monetary policy

The standard tool is the policy interest rate — Fed funds rate, ECB deposit facility rate, BoE Bank Rate, BoJ short-term policy rate. The central bank steers the overnight interbank rate via:

Open market operations (OMO): purchases/sales of short-term government securities.
Standing facilities: lending facility (discount window/marginal lending) and deposit facility (interest on reserves, IORB at the Fed since 2008).
Reserve requirements (largely defunct in most advanced economies; the Fed set them to zero in March 2020).
Repo and reverse repo operations: the Fed’s overnight reverse repo (ON RRP) facility was created in 2013 and absorbed ~$2.5 trillion at its 2022–23 peak.

Since 2008 most advanced-economy central banks operate in an ample-reserves regime (Fed terminology) — the Fed sets IORB and ON RRP as the floor and ceiling of a rate corridor, and overnight markets settle within. The 2019 repo spike and 2023 SVB episode both prompted reconsideration of optimal reserve levels; the Fed’s 2024–25 balance-sheet-runoff slowdown reflects this.

Unconventional monetary policy at the zero lower bound (ZLB)

When policy rates hit zero (or in some cases negative — ECB, SNB, BoJ in 2014–22), central banks deploy:

Quantitative easing (QE): large-scale asset purchases (LSAPs). Fed QE1 (2008–10, $1.7 TMBS + T re a s u r i es), QE 2 (2010-11,$ 600B Treasuries), QE3 (2012–14, open-ended), QE-Infinity (2020–22, $4.6 T), t h e n q u an t i t a t i v e t i g h t e nin g (QT) f ro m J u n e 2022. T h e F e d ba l an ces h ee tp e ak e d a t$ 8.95T in April 2022 and was $6.8T as of May 2026. Bernanke’s 2022 Nobel lecture argued QE works mainly through portfolio rebalancing and signaling, less through pure duration absorption.
Yield curve control (YCC): BoJ’s flagship policy from September 2016 to March 2024, targeting the 10-year JGB yield (initially at 0%, widened to ±0.25% in 2022, ±0.5% in late 2022, ±1.0% in 2023, then abandoned in March 2024 alongside the end of negative rates). The RBA briefly tried YCC at the 3-year point (2020–21) with mixed results.
Forward guidance: communication about the future path of policy rates. Two flavors — Odyssean (commitment, e.g., ECB’s 2013 “rates at or below current levels for an extended period”) and Delphic (forecast, e.g., the Fed’s Summary of Economic Projections dot plot since 2012). Eggertsson and Woodford (2003) provided the theoretical case.
Negative interest rates (NIRP): ECB deposit rate −0.5% (2019–22), SNB −0.75% (2015–22), BoJ −0.1% (2016–24), Riksbank −0.5% (2015–19). Theoretical limit (the “reversal rate”, Brunnermeier and Koby 2018) is set by bank profitability and physical cash storage costs.
Funding-for-lending schemes: ECB TLTROs (I-IV), BoE Term Funding Scheme, BoJ Loan Support Program — subsidized lending to incentivize bank credit.
Direct asset purchases beyond government bonds: ECB Corporate Sector Purchase Programme (2016–), Fed corporate bond facility (2020, briefly), BoJ ETF and J-REIT purchases (BoJ became the largest single owner of Japanese equities at peak, ~7% of TOPIX).

Central bank digital currencies (CBDC)

As of 2026, three major CBDCs are live (e-CNY in China — $250B+ in transactions cumulatively, eNaira in Nigeria, Bahamian Sand Dollar) and three in pilot (ECB’s digital euro, expected launch decision 2026 Q3; Brazil’s Drex; India’s e-rupee). The Fed has explicitly declined to launch a retail CBDC; FedNow (instant payments, launched July 2023) is the U.S. alternative. The BIS Innovation Hub’s Project Agorá (multi-CBDC cross-border platform with Fed, ECB, BoE, BoJ, BoK, Banco de México, SNB, MAS) entered its second phase in early 2026.

8. Fiscal Policy

Multipliers

The fiscal multiplier is the ratio of the change in output to the change in autonomous fiscal spending. In the simple Keynesian cross with marginal propensity to consume c, the multiplier is 1/(1−c). Empirical estimates vary widely by:

State of the economy: larger in recessions (Auerbach and Gorodnichenko 2012 estimate ~2.5 in recession vs ~0.5 in expansion).
Monetary regime: larger at the ZLB (Christiano, Eichenbaum, and Rebelo 2011 find ~3 at the ZLB).
Spending type: government investment multipliers ~1.0–1.5; transfer multipliers ~0.5–1.0; tax-cut multipliers depend on permanence and targeting (Romer and Romer 2010).
Trade openness: smaller in open economies due to import leakage.

Ramey (2019, Handbook of Macroeconomics) summarizes the literature; the consensus point estimate for a closed-economy non-ZLB advanced economy is in the range 0.6–1.2.

Debt sustainability

The government’s intertemporal budget constraint requires that the present value of future primary surpluses equal current debt. The dynamic debt-to-GDP equation:

Δ(b) ≈ (r − g)·b − ps

where b is debt/GDP, r is the real interest rate on government debt, g is real GDP growth, and ps is the primary surplus/GDP. When r < g (the “r minus g” regime that prevailed in much of the 2010s), debt is self-stabilizing even with primary deficits — the basis of Blanchard’s (2019) Presidential Address arguing that public debt may have no fiscal cost in the low-rate environment.

The 2022–24 rate-hiking cycle reversed r − g for the U.S. (10y TIPS ~2.0% vs trend real growth ~1.8%), revived debt-sustainability concerns, and prompted the IMF’s October 2024 Fiscal Monitor warning on U.S. and U.K. trajectories. CBO’s January 2026 Long-Term Budget Outlook projects U.S. federal debt held by the public reaching 156% of GDP by 2055 under current law, up from 99% in 2026.

Fiscal rules

EU Stability and Growth Pact (1997, revised 2024): 3% deficit, 60% debt thresholds; the 2024 reform replaced the 1/20 debt-reduction rule with country-specific medium-term plans.
U.S. statutory PAYGO (2010): largely non-binding; the debt ceiling is the operative constraint, with the latest crisis resolved by the Fiscal Responsibility Act of June 2023.
U.K. fiscal rules: revised under each Chancellor; the 2024 Labour government adopted a “stability rule” (current budget balance by year five) and an “investment rule” (PSND ex-BoE falling by year five).
Switzerland’s debt brake (2003): the gold standard; a constitutional structural-balance rule that has reduced gross debt from 50% to 17% of GDP.

9. Growth Theory

Solow-Swan model

Solow (1956, Nobel 1987) and Swan (1956): Y = F(K, AL) with constant returns. Per effective worker, y = f(k), and capital accumulation Δk = sf(k) − (n + g + δ)k converges to a steady state where ksatisfies sf(k) = (n + g + δ)k*. Predictions:

Conditional convergence: countries with similar parameters converge; absolute convergence fails (Mankiw, Romer, Weil 1992 add human capital and rescue conditional convergence).
No long-run growth from saving: a higher saving rate raises the level of output per capita but not its long-run growth rate, which is g (exogenous TFP growth) in the Solow model.
Golden rule: consumption-maximizing k* satisfies f’(k_gold) = n + g + δ.

The Solow residual — TFP growth as a residual after netting out capital and labor contributions — is the workhorse of growth accounting (Jorgenson, Gollop, and Fraumeni 1987; Penn World Table). It includes everything other than measured inputs: technology, allocative efficiency, institutional quality, measurement error.

Endogenous growth

Romer (1986, 1990, Nobel 2018): nonrival ideas + monopolistic competition + intentional R&D investment produces endogenous growth. Output Y = A·K^α(HL)^{1−α}; A grows via δA/A = δ·H_A/A, where H_A is human capital devoted to research. The model predicts:

Scale effects: bigger economies grow faster (criticized empirically; Jones 1995 modifies to “semi-endogenous”).
Knowledge spillovers: justify R&D subsidies and patent policy.

Aghion and Howitt (1992) introduced Schumpeterian “creative destruction”: new innovators displace incumbent monopolists. The model rationalizes both growth and obsolescence and has been the workhorse of innovation economics. Aghion, Akcigit, and Howitt (2014) update the framework with firm-level data; Acemoglu and Restrepo (2020) extend to automation and tasks.

Long-run growth puzzles

The Great Stagnation (Cowen 2011, Gordon 2016): U.S. TFP growth slowed from ~1.8% (1947–2007) to ~0.5% (2007–2019). Hypotheses: exhaustion of post-war ideas, mismeasurement of digital goods, declining R&D productivity (Bloom, Jones, Van Reenen, Webb 2020: “Are Ideas Getting Harder to Find?”), market power, demographics.
The AI productivity question (Brynjolfsson and McAfee 2014; Brynjolfsson, Rock, Syverson 2021 “productivity J-curve”; Acemoglu 2024 “The Simple Macroeconomics of AI”): does generative AI produce a TFP renaissance? Brynjolfsson’s CodeNet, customer-service, and writing experiments (2023–25) show 14–40% productivity gains for low-skill workers. Acemoglu’s 2024 calibration estimates only +0.5% cumulative TFP over 10 years from current AI; Goldman Sachs and McKinsey project +1.5%/year. The data through 2026 Q1 show U.S. nonfarm-business labor productivity growth of ~2.3% (2023–25 average), well above the 2007–2019 trend but consistent with both cyclical recovery and an early AI effect — the literature remains unresolved.

10. Labor Markets

Stocks and flows

The labor force = employed + unemployed (actively searching). Unemployment rate U/(E+U). The U.S. BLS publishes six measures: U-1 (long-term unemployed) through U-6 (broadest, including marginally attached and part-time for economic reasons). U-3 is the headline rate.

Labor force participation rate (LFPR): (E+U)/working-age population. The U.S. prime-age (25–54) LFPR was 83.6% in April 2026 — slightly above the pre-COVID peak. The overall LFPR (62.6%) remains below pre-COVID (63.4%) due to demographic aging.

JOLTS and the Beveridge curve

The BLS Job Openings and Labor Turnover Survey (JOLTS, monthly since 2000) tracks job openings (V), hires, quits, and layoffs. The quits rate is a coincident indicator of labor-market tightness — the “Great Resignation” of 2021–22 peaked at 3.0% (a record).

The Beveridge curve plots unemployment u against vacancies v: a downward-sloping locus where matching efficiency, sectoral reallocation, and search frictions show up as shifts. Diamond, Mortensen, and Pissarides (joint Nobel 2010) provided the search-and-matching foundation. Blanchard, Domash, and Summers (2022) used Beveridge-curve shifts to argue the post-COVID labor market was uniquely tight; Waller’s 2022 speech (Fed) used the same diagram to argue a soft landing was achievable via vacancy declines without u rising — which is roughly what played out 2023–25 (vacancies fell from 12 million in March 2022 to 7.4 million in April 2026; U-3 rose only from 3.4% to 4.1%).

Wage Phillips curve and the wage-price spiral

The wage Phillips curve relates nominal wage growth to unemployment plus inflation expectations. Gagnon and Sarsenbayev (2022) and Bernanke-Blanchard (2023) find limited evidence of a wage-price spiral in the 2021–23 episode — wages chased prices rather than leading them, and the dynamic dampened as supply shocks reversed. The IMF’s October 2022 World Economic Outlook chapter on wage-price spirals reached the same conclusion using cross-country data.

11. International Economics: Exchange Rates and the Balance of Payments

The balance of payments

BoP = Current Account + Capital Account + Financial Account + Errors & Omissions = 0 (identity).

Current account: trade balance (goods + services) + primary income (investment income) + secondary income (transfers). U.S. CA deficit ~3.0% of GDP through 2025.
Capital account: small; debt forgiveness, migrant transfers.
Financial account: FDI, portfolio investment, other investment (loans, deposits), reserve assets.

A current account deficit must be financed by a financial account surplus — capital inflows. The U.S. “exorbitant privilege” (Eichengreen 2011) is the ability to run persistent CA deficits while earning a positive net investment income return — because U.S. liabilities are low-yielding safe assets (Treasuries, dollar deposits) while U.S. assets abroad are higher-yielding FDI and equity. Gourinchas and Rey (2007) document this.

Exchange rate regimes

The IMF classifies regimes on a spectrum: hard pegs (currency board, dollarization) → soft pegs (conventional, crawling, basket) → floating (managed, free). 2024 distribution: ~12% hard pegs, ~35% soft pegs, ~35% floating, ~18% other.

Bretton Woods (1944–73): par-value system with U.S. dollar tied to gold at $35/oz and other currencies pegged to the dollar. Triffin (1960) identified the Triffin dilemma: the system required U.S. deficits to supply dollar liquidity, but accumulating dollar liabilities eventually undermined the gold peg — which Nixon ended in August 1971. The modern variant: dollar reserve hegemony requires U.S. deficits but creates long-run sustainability questions about the dollar’s role.

Euro area (1999–): a unified currency zone with one monetary policy, fragmented fiscal policy, and incomplete banking and capital-markets union. Mundell’s (1961) optimal currency area criteria — factor mobility, fiscal transfers, business-cycle synchronization, price/wage flexibility — are imperfectly satisfied. The 2010–12 sovereign debt crisis exposed the design flaw; Draghi’s “whatever it takes” (July 2012) and OMT averted breakup. Post-2020 NGEU (NextGenerationEU) and SURE introduced the first joint debt issuance, a step toward fiscal union; the 2024 reform of the Stability and Growth Pact and the 2025 capital markets union initiative continue the integration.

Exchange rate theories

Purchasing power parity (PPP): long-run anchor, fails badly short-run. Big Mac Index (The Economist) and OECD PPPs are standard measurements.
Uncovered interest parity (UIP): E[ΔS] = i − i*. Famously fails — the “forward premium puzzle” (Fama 1984): high-yield currencies appreciate on average, the basis of the carry trade. Engel (2014, Handbook of International Economics) surveys.
Real exchange rate: q = SP*/P. Long-run mean-reverting but with half-lives of 3–5 years (the PPP puzzle of Rogoff 1996).
Behavioral and microstructure approaches: order flow (Evans and Lyons 2002), scapegoat (Bacchetta-van Wincoop 2004), rare disasters.

Dollar dominance and de-dollarization

As of 2026 Q1:

USD share of global FX reserves (IMF COFER): 57.4% (down from 71% in 2000).
USD share of cross-border bank claims (BIS): 50%.
USD share of trade invoicing: ~50% (much higher outside intra-EU and intra-Asia trade).
USD share of SWIFT messaging: 49.0% (CNY 4.3%, EUR 22%, GBP 7%, JPY 4%).

The BRICS+ (Brazil, Russia, India, China, South Africa, plus 2024 expansions UAE, Iran, Egypt, Ethiopia) discussions of a common settlement unit (informally “R5” or “BRICS Pay”) have produced bilateral CNY-RUB and CNY-INR trade settlement but no functioning supranational currency. Eichengreen (2024) and Pozsar (2022) survey the trajectory.

12. Inflation: The 2021–2025 Cycle

Anatomy of the surge

U.S. CPI headline year-over-year inflation:

Period	Headline CPI	Core CPI	Drivers
2020 Avg	1.2%	1.7%	COVID disinflation, services collapse
2021 Q4	6.8%	5.0%	Demand surge, goods/durable bottleneck, base effects
2022 Jun (peak)	9.1%	5.9%	Russia-Ukraine, energy, food, rent acceleration
2023 Q4	3.4%	4.0%	Supply chains heal, goods deflation, services sticky
2024 Q4	2.9%	3.3%	Rent disinflation, shelter lag resolves
2025 Q4	2.6%	2.7%	Near target, services normalization
2026 Apr	2.4%	2.5%	Approaching 2% target; Fed cutting since Sept 2024

The euro area HICP peaked at 10.6% in October 2022 (Russia-Ukraine gas shock dominant); the U.K. CPI peaked at 11.1% in October 2022 (energy + sterling); Japan finally exited multi-decade deflation, hitting 4.3% in January 2023 and stabilizing near 2% by 2025.

Decompositions

San Francisco Fed Cyclical vs. Acyclical CPI (Mahedy and Shapiro 2017, ongoing): both components rose in 2021–22, but cyclical accounted for the majority of the persistence in 2023–24.
Bernanke-Blanchard (2023) decomposition for the U.S., euro area, U.K., Japan: 75–85% of the 2021–23 surge was supply shocks (energy, food, supply chains); the remainder was tight labor markets via the wage Phillips curve. Expectations remained well-anchored throughout.
NY Fed Underlying Inflation Gauge (UIG): peaked at 4.8% in 2022, back to 2.7% by April 2026 — a slow normalization.
PCE Sticky-vs-Flexible (Atlanta Fed Bryan-Meyer): sticky CPI is back to 2.5% (2026 Q1), flexible to 1.0% — composition matters for which signal the Fed reads.

Policy responses

The Fed lifted the funds rate target from 0–0.25% (March 2022) to 5.25–5.50% (July 2023 – September 2024), held it for 14 months, then cut to 4.50–4.75% (September 2024), 4.25–4.50% (December 2024), 3.75–4.00% (March 2025), and as of May 2026 stands at 3.25–3.50%. The Fed’s official 2026 view (March 2026 SEP) projects two more cuts to 2.75–3.00% by year-end and a neutral rate near 3.0% (up from the 2.5% estimate of the late 2010s — the post-pandemic re-rating of r*).

ECB deposit rate: 2.0% (May 2026), down from 4.0% peak (Sept 2023–Jun 2024). BoE Bank Rate: 3.75% (May 2026), down from 5.25% peak. BoJ: 0.50% policy rate (March 2026 hike), exiting NIRP+YCC.

Productivity surprise

A key feature of 2023–25 was that disinflation occurred without the recession most forecasters expected (the “immaculate disinflation”). Bernanke and Blanchard (2023, updated 2025) attribute this to:

Supply-shock reversal (energy, supply chains) doing most of the work.
Anchored expectations preventing a wage-price spiral.
Productivity acceleration (some attribute to AI, but the time series is too short for confidence).
Immigration-driven labor supply expansion (U.S. labor force grew 3.5 million 2022–24 from net immigration, easing wage pressure — CBO January 2024 revisions).

13. Financial Crises

Diamond-Dybvig and the theory of bank runs

Diamond and Dybvig (1983, joint 2022 Nobel with Bernanke): banks transform illiquid long-term assets into liquid short-term liabilities (demand deposits). This is socially valuable but vulnerable to self-fulfilling runs — a bad equilibrium exists where each depositor’s belief that others will withdraw makes withdrawal optimal. Solutions: deposit insurance (FDIC 1933, removes the run incentive); suspension of convertibility; lender of last resort (Bagehot 1873: lend freely at a penalty rate against good collateral).

Bernanke’s 1983 AER paper showed empirically that bank failures during the Great Depression amplified the contraction via disrupted credit intermediation — the credit channel of monetary transmission. This empirical foundation underwrites the Fed’s aggressive 2008 and 2020 liquidity responses.

The 2008 Global Financial Crisis

Triggered by the U.S. subprime mortgage bust (2006–07), the crisis became systemic in September 2008 (Lehman bankruptcy September 15, AIG rescue September 16, money market fund Reserve Primary “breaking the buck” September 17). The Fed’s response, led by Bernanke (then Chair, drawing on his academic work):

Emergency lending facilities: TAF, TSLF, PDCF, AMLF, CPFF, MMIFF, TALF — broad expansion of LOLR to non-banks under Section 13(3).
Aggressive rate cuts: Fed funds from 5.25% (Sept 2007) to 0–0.25% (Dec 2008).
QE1 ($1.7T, MBS + agency + long Treasuries).
Stress testing: SCAP 2009, then annual CCAR/DFAST.
TARP ( $700 BT re a s u ry a u t h or i t y, o f w hi c h$ 450B disbursed; final 2014 net profit to Treasury).
Dodd-Frank (July 2010): Volcker Rule, SIFI designation, OLA resolution authority, derivatives clearing, CFPB, FSOC.

Reinhart and Rogoff’s (2009) This Time Is Different documented systematic patterns: pre-crisis credit booms, post-crisis output losses of 9% (median), unemployment +7pp, real housing −36%, real equity −56%, fiscal debt +86% over three years. The U.S. tracked these benchmarks closely.

2020 COVID liquidity crisis

Mid-March 2020: the dash for cash. Treasury market dysfunction — the safest asset in the world saw bid-ask spreads widen 10x. Causes: hedge fund cash-futures basis trade unwind, foreign official selling, mutual fund redemptions, primary dealer balance sheet capacity. The Fed’s response was unprecedented in speed and scope:

Rates to 0–0.25% (March 15).
QE-Infinity announced March 23 ( $75 B / d a y T re a s u r i es +$ 50B/day MBS).
Reopened 13(3) facilities (CPFF, MMLF, PDCF) plus new ones (PMCCF, SMCCF, MLF, MSLP, TALF 2.0).
Coordinated central bank swap lines expanded (5 → 14 lines).
$4.6T of asset purchases over 2020–22.

Treasury under the CARES Act ($2.2T) provided fiscal backstops to Fed facilities and direct fiscal stimulus. The Treasury market stabilized within 10 trading days — the fastest crisis response in central bank history. Logan (Dallas Fed President, formerly NY Fed Markets Group), Liang (Treasury Domestic Finance, formerly Fed), and Duffie (Stanford) have argued for permanent structural reforms (Treasury market clearing, standing repo facility, regulatory relief for dealer balance sheets). The SEC’s 2024 Treasury clearing rule (mandatory CCP clearing of repo from June 2026, cash from December 2026) is the main reform outcome.

2022 U.K. gilt crisis

September 23, 2022: U.K. Chancellor Kwasi Kwarteng’s “mini-Budget” announced £45B unfunded tax cuts. Within five trading days, 30-year gilt yields rose 130bps, the largest move on record. The crisis was amplified by liability-driven investment (LDI) strategies at U.K. defined-benefit pension funds: levered long-duration positions hedging pension liabilities faced margin calls, forced selling, and risk of insolvency. The BoE intervened September 28 with up to £65B of long-dated gilt purchases — a financial-stability operation framed as outside the MPC’s monetary remit. Truss resigned October 20, 44 days into her premiership.

The episode revealed (i) non-bank financial intermediary (NBFI) vulnerabilities at scale, (ii) the limits of fiscal sovereignty when bond markets revolt, and (iii) the tension between QT (the BoE had just started actively selling gilts) and financial-stability buyer-of-last-resort operations.

2023 SVB / regional bank crisis

March 10, 2023: Silicon Valley Bank (16th-largest U.S. bank, $209 B a sse t s) f ai l e d a f t er a$ 42B run in 24 hours — the fastest in U.S. history. Causes: concentrated tech-startup deposit base, large held-to-maturity Treasury and MBS portfolio with unrealized losses (~ $15 B a t e n d - 2022, e x cee d in g t an g ib l ee q u i t y), re gu l a t oryro ll ba c k o f EPSt h res h o l d s f orc a t e g ory I Vbank s in 2018. S i g na t u re B ank f ai l e d M a rc h 12 ($ 110B). First Republic failed May 1 ($229B; sold to JPMorgan).

Fed response:

Bank Term Funding Program (BTFP): new facility lending against Treasuries and MBS at par (rather than market) value, one-year maturity. Peaked at $165B; closed March 2024.
Systemic risk exception invoked for SVB and Signature, protecting all (including uninsured) depositors.
Treasury, FDIC, Fed joint statement March 12 backstopping the system.
Discount window borrowing surged to $153B (week of March 15) — the highest since 2008.

Aftermath: Barr (Fed Vice Chair for Supervision) “Holistic Capital Review” 2023; final Basel III “endgame” rules issued by federal regulators in revised form September 2024 (after industry pushback on the July 2023 proposal); FDIC special assessment to recapitalize the Deposit Insurance Fund.

Other recent episodes worth noting

China Evergrande default (December 2021): largest emerging-market dollar bond default; restructuring still ongoing. The property sector contraction continues to weigh on PBoC policy through 2026.
Turkey lira crisis (2018, 2021, 2023): unorthodox monetary policy (Erdoğan’s “interest rates cause inflation” doctrine) produced repeated currency crises. The May 2023 election aftermath saw a return to orthodox policy under Şimşek; policy rate hiked from 8.5% to 50% within a year.
Sri Lanka sovereign default (April 2022): first South Asian default in modern era; reconstructed under IMF EFF.
Argentina (perpetual): Milei’s December 2023 election produced a fiscal-shock-therapy program; 2024–25 inflation fell from 211% to 117%, primary surplus achieved 2024.

14. AI, Productivity, and the 2024–26 Debate

The biggest open question in 2026 macroeconomics: is generative AI producing measurable TFP gains?

Brynjolfsson camp (optimist)

Brynjolfsson, Li, Raymond (2023, QJE) field experiment with a Fortune 500 customer-service software firm: AI assistance raised resolved-issues-per-hour by 14% on average, 35% for least-experienced workers. The first credible RCT showing AI productivity effects in real production.
Noy and Zhang (2023) writing experiment: 37% time reduction with ChatGPT.
Peng et al. (2023) GitHub Copilot experiment: 56% completion-time reduction.
Brynjolfsson and McAfee (The Second Machine Age, 2014; Power and Prediction with Agrawal-Gans-Goldfarb 2022): general purpose technology, productivity J-curve (lag then surge).

Acemoglu camp (skeptic)

Acemoglu (2024) NBER WP 32487 “The Simple Macroeconomics of AI”: back-of-envelope using task-share calculations and conservative productivity gains suggests only +0.5% TFP cumulative over 10 years from current AI.
Acemoglu and Restrepo (2018, 2020, 2022) automation-and-tasks framework: AI as a “so-so” automation technology may displace labor without raising productivity proportionally.
Acemoglu and Johnson (Power and Progress, 2023): historically, technology adoption is shaped by political-economy choices; productivity gains don’t automatically translate to broad welfare.

The data through 2026 Q1

U.S. nonfarm-business labor productivity growth (BLS): 2023 +2.5%, 2024 +2.7%, 2025 +2.4%, 2026 Q1 +2.6% (annualized). The 2023–25 average of 2.5% is well above the 2007–2019 trend of 1.4%.
TFP growth (Fernald Fed SF measure, utilization-adjusted): 2023 +1.4%, 2024 +1.5%, 2025 +1.3%. Above trend (~0.6% 2007–2019) but only modestly.
Investment in IP products (software, R&D) accelerating; AI-specific capex by hyperscalers (Microsoft, Alphabet, Amazon, Meta) reached $260 B in 2025 an d i s p ro j ec t e d$ 310B in 2026.

The signal is real but its persistence and scope are not yet settled. The conventional wisdom (Bernanke-Sahm-Summers panel at Brookings, April 2026) is that the AI productivity effect is plausibly real but the 2023–25 acceleration also reflects post-pandemic reallocation gains (Foster-Haltiwanger-Krizan dynamics) and that 2027–30 will be the diagnostic period.

15. Industrial Policy Renaissance

The 2020s saw a return of state-directed industrial policy in advanced economies, breaking with the post-1980 consensus.

U.S. CHIPS and Science Act (August 2022): $52.7B for semiconductor manufacturing and R&D, 25% investment tax credit. Awards to TSMC (Arizona fabs), Intel (Ohio, Arizona), Samsung (Texas), Micron (New York). As of March 2026,$ 33B of grants finalized, $25 B o f l o an s,$ 200B+ of induced private capex announced.
U.S. Inflation Reduction Act (August 2022): ~ $370 B (or i g ina ll y) o f c l ima t e / e n er g y t a x cre d i t s, u n c a pp e d . J o in tC o mmi tt eeo n T a x a t i o n re v i se d t h e 10 - ye a res t ima t e t o$ 870B (March 2024) due to higher uptake. The 45X manufacturing credit and 45V hydrogen credit drove large announced investments; the change in administration in 2025 introduced policy uncertainty about which credits survive — the 45X production credit and 45Y/48E technology-neutral credits are politically resilient; 45V faces possible curtailment in 2026 reconciliation.
U.S. Bipartisan Infrastructure Law (November 2021): $1.2 T o f w hi c h$ 550B new spending; infrastructure, broadband, EV charging, grid.
EU Green Deal Industrial Plan and Net-Zero Industry Act (March 2023): target of 40% of clean-tech needs met domestically by 2030; faster permitting; €270B+ Innovation Fund and InvestEU.
EU Chips Act (September 2023): €43B mobilized for semiconductors; target of 20% global market share by 2030.
Japan METI semiconductor strategy: subsidies to TSMC (Kumamoto), Rapidus (Hokkaido 2nm by 2027), Micron (Hiroshima).
South Korea K-Chips Act (2023): tax credits to Samsung and SK Hynix.
China dual circulation and “new productive forces” (Xi 2023–25 framing): advanced manufacturing, EVs, batteries, solar, AI; production overcapacity in EVs/solar/batteries triggering trade frictions with EU and U.S.

The macro implications: higher public investment, potentially higher trend growth (if effective), inflation pressure in supply-constrained sectors (semiconductor labor, power equipment), and a re-fragmentation of global supply chains (Friend-shoring, near-shoring). Aghion, Cherif, and Hasanov (2024) and Rodrik (2023) survey the new industrial-policy literature; the IMF’s October 2024 WEO chapter on industrial policy provides cross-country evidence.

16. Tools, Datasets, and Workflow

Datasets and data services

FRED (Federal Reserve Bank of St. Louis): the single most useful open macro data API. 800,000+ series. Python: fredapi; R: fredr.
Haver Analytics: paid premium aggregator used by most central banks and major dealers.
Bloomberg Terminal: real-time markets + ECON/WECO/CGEM/IBQ functions for macro data; gold standard for sell-side.
OECD.Stat / OECD Data Explorer: harmonized cross-country statistics; OECD Economic Outlook database.
IMF WEO (World Economic Outlook) database: country-level annual data and forecasts; IMF DataMapper API.
IMF IFS (International Financial Statistics): balance of payments, monetary, exchange rates.
BIS Statistics: cross-border banking, FX turnover (Triennial Survey), credit-to-GDP gaps.
World Bank WDI (World Development Indicators): development-oriented panel.
Penn World Table 10.01 (Groningen): cross-country productivity and price levels.
Maddison Project Database: historical GDP per capita 1–2018.
AMECO (European Commission Annual Macro-Economic Database): EU-focused.
BEA NIPA tables: U.S. national accounts at source.
BLS CES, CPS, JOLTS, CPI, PPI: U.S. labor-market and price data.
Fed H.4.1, H.6, H.8, H.15: balance sheet, money stock, bank credit, interest rates.
NY Fed Survey of Consumer Expectations (SCE), Michigan Survey of Consumers: inflation and labor expectations.
Atlanta Fed GDPNow, NY Fed Nowcast: real-time nowcasting.
EIA STEO, IEA Oil Market Report: energy.

Software stack

R: tidyverse, tsibble, fable, vars, dynlm, tseries, forecast, rstan for Bayesian; dynare and gEcon for DSGE.
Python: pandas, statsmodels, linearmodels, arch (volatility), prophet, pymc (Bayesian), dolo and dolark (DSGE/HANK), quantecon (Stachurski-Sargent codebase).
MATLAB: still standard for DSGE estimation via Dynare (Adjemian et al.) — the dominant DSGE estimation toolbox; IRIS (Plesha) alternative; OccBin (Guerrieri-Iacoviello) for occasionally binding constraints; Dynare++ for higher-order perturbation.
Julia: Distributions.jl, Turing.jl, DSGE.jl (NY Fed open-source DSGE codebase), DifferentialEquations.jl for continuous-time HANK (Achdou-Han-Lasry-Lions-Moll).
Stata: still ubiquitous for empirical macro (event studies, local projections — Jordà 2005).

17. Cross-References

This note pairs with:

microeconomics-foundations — household and firm optimization; the micro-foundations that underpin DSGE and HANK.
corporate-finance-and-markets — asset pricing, term structure, credit spreads; the financial-market side of the macro-finance literature (Brunnermeier, He-Krishnamurthy, Adrian-Crump-Moench).
electricity-markets — the energy supply shock channel that dominated 2021–22 inflation; carbon pricing and the macro of the energy transition.
_index — climate-macro literature (Nordhaus DICE, Stern Review, NGFS scenarios, transition risk and stranded assets).
hypothesis-testing-mle — the maximum-likelihood and Bayesian estimation machinery used in DSGE, BVAR, and macro-finance models.

Related sub-topics to expand in future notes:

Heterogeneous-agent macro: HANK, TANK, two-asset HANK, perpetual-youth life-cycle, sufficient-statistic approaches (Kaplan-Moll-Violante, Auclert, Achdou-Han-Lasry-Lions-Moll).
Macro-finance: financial accelerator, banking in DSGE, intermediary asset pricing, macroprudential policy.
Open-economy: SOE-DSGE, exchange-rate disconnect, original sin redux (Carstens-Shin), dominant currency paradigm (Gopinath-Itskhoki-Maggiori-Rey).
New monetarist and search-theoretic monetary economics (Lagos-Wright, Williamson-Wright).
Climate macro: integrated assessment, climate-DSGE (Golosov-Hassler-Krusell-Tsyvinski), transition risk in central bank scenarios (NGFS).
Distributional national accounts and inequality (Piketty, Saez, Zucman; Auten-Splinter critique).

18. Citations and Nobel Laureates

A working bibliography organized by topic; this is the canon that any macroeconomist should know.

Foundational pre-1970 figures

Keynes, J.M. (1936). The General Theory of Employment, Interest and Money. Macmillan.
Hicks, J. (1937). “Mr. Keynes and the Classics: A Suggested Interpretation”. Econometrica 5(2). Nobel 1972 (with Arrow).
Friedman, M. (1957). A Theory of the Consumption Function. NBER.
Solow, R. (1956). “A Contribution to the Theory of Economic Growth”. QJE 70(1). Nobel 1987.
Tobin, J. (1958). “Liquidity Preference as Behavior Towards Risk”. Review of Economic Studies 25. Nobel 1981.
Phillips, A.W. (1958). “The Relation between Unemployment and the Rate of Change of Money Wage Rates in the United Kingdom, 1861–1957”. Economica 25.
Mundell, R. (1961). “A Theory of Optimum Currency Areas”. AER 51(4). Nobel 1999.
Mundell, R. (1963). “Capital Mobility and Stabilization Policy under Fixed and Flexible Exchange Rates”. Canadian Journal of Economics 29(4).
Triffin, R. (1960). Gold and the Dollar Crisis. Yale.

The monetarist and rational-expectations revolutions

Friedman, M. (1968). “The Role of Monetary Policy”. AER 58(1). Nobel 1976 “for his achievements in the fields of consumption analysis, monetary history and theory and for his demonstration of the complexity of stabilization policy”.
Phelps, E. (1967, 1968). “Phillips Curves, Expectations of Inflation and Optimal Unemployment over Time”. Economica. Nobel 2006 “for his analysis of intertemporal tradeoffs in macroeconomic policy”.
Lucas, R. (1972). “Expectations and the Neutrality of Money”. JET 4. Nobel 1995.
Lucas, R. (1976). “Econometric Policy Evaluation: A Critique”. Carnegie-Rochester.
Sargent, T. and Wallace, N. (1975). “Rational Expectations, the Optimal Monetary Instrument, and the Optimal Money Supply Rule”. JPE. Sargent Nobel 2011 with Sims (separately for SVAR — Sims 1980, Econometrica 48).
Kydland, F. and Prescott, E. (1977). “Rules Rather than Discretion: The Inconsistency of Optimal Plans”. JPE. Joint Nobel 2004.
Kydland, F. and Prescott, E. (1982). “Time to Build and Aggregate Fluctuations”. Econometrica — the founding RBC paper.

New Keynesian and modern macro

Calvo, G. (1983). “Staggered Prices in a Utility-Maximizing Framework”. JME 12.
Taylor, J. (1993). “Discretion versus Policy Rules in Practice”. Carnegie-Rochester.
Clarida, R., Galí, J., and Gertler, M. (1999). “The Science of Monetary Policy: A New Keynesian Perspective”. JEL 37(4).
Woodford, M. (2003). Interest and Prices. Princeton.
Galí, J. (2008, 2015). Monetary Policy, Inflation, and the Business Cycle. Princeton.
Smets, F. and Wouters, R. (2003, 2007). “An Estimated Stochastic Dynamic General Equilibrium Model of the Euro Area” / “Shocks and Frictions in US Business Cycles”. JEEA / AER.
Kaplan, G., Moll, B., and Violante, G. (2018). “Monetary Policy According to HANK”. AER 108(3).

Growth theory

Romer, P. (1986, 1990). “Increasing Returns and Long-Run Growth” / “Endogenous Technological Change”. JPE. Nobel 2018 with Nordhaus.
Aghion, P. and Howitt, P. (1992). “A Model of Growth Through Creative Destruction”. Econometrica 60(2).
Mankiw, N.G., Romer, D., and Weil, D. (1992). “A Contribution to the Empirics of Economic Growth”. QJE 107(2).
Acemoglu, D. (2009). Introduction to Modern Economic Growth. Princeton.
Acemoglu, D. and Restrepo, P. (2018, 2020). “The Race Between Man and Machine” / “Robots and Jobs”. AER / JPE.
Bloom, N., Jones, C., Van Reenen, J., Webb, M. (2020). “Are Ideas Getting Harder to Find?“. AER 110(4).

International and exchange rates

Dornbusch, R. (1976). “Expectations and Exchange Rate Dynamics”. JPE 84(6) — the overshooting model.
Obstfeld, M. and Rogoff, K. (1995). “Exchange Rate Dynamics Redux”. JPE 103(3).
Krugman, P. (1979). “A Model of Balance-of-Payments Crises”. JMCB 11. Nobel 2008 “for his analysis of trade patterns and location of economic activity”.
Krugman, P. (1980). “Scale Economies, Product Differentiation, and the Pattern of Trade”. AER 70(5).
Gopinath, G., Boz, E., Casas, C., Díez, F., Gourinchas, P.O., Plagborg-Møller, M. (2020). “Dominant Currency Paradigm”. AER 110(3).
Gourinchas, P.O. and Rey, H. (2007). “From World Banker to World Venture Capitalist”. in G7 Current Account Imbalances, NBER.

Financial crises and banking

Bernanke, B. (1983). “Nonmonetary Effects of the Financial Crisis in the Propagation of the Great Depression”. AER 73(3). Nobel 2022 (jointly with Diamond and Dybvig) “for research on banks and financial crises”.
Diamond, D. and Dybvig, P. (1983). “Bank Runs, Deposit Insurance, and Liquidity”. JPE 91(3). Joint Nobel 2022.
Bernanke, B., Gertler, M., and Gilchrist, S. (1999). “The Financial Accelerator in a Quantitative Business Cycle Framework”. Handbook of Macroeconomics.
Reinhart, C. and Rogoff, K. (2009). This Time Is Different. Princeton.
Brunnermeier, M. and Sannikov, Y. (2014). “A Macroeconomic Model with a Financial Sector”. AER 104(2).
Gertler, M. and Karadi, P. (2011). “A Model of Unconventional Monetary Policy”. JME 58(1).
Diamond, P., Mortensen, D., and Pissarides, C. — joint Nobel 2010 for search and matching theory of labor markets.

Recent policy macro (2020–2025)

Bernanke, B. and Blanchard, O. (2023, updated 2025). “What Caused the U.S. Pandemic-Era Inflation?“. Brookings Hutchins Center / Peterson IIE.
Bernanke, B. (2022). Nobel Prize Lecture: “Banking, Credit, and Economic Fluctuations”.
Blanchard, O. (2019). “Public Debt and Low Interest Rates”. AER 109(4) — Presidential Address to the AEA.
Blanchard, O. (2023). Fiscal Policy under Low Interest Rates. MIT Press.
Powell, J. (annual). Jackson Hole speech, Kansas City Fed Economic Symposium — the most-watched Fed communication.
Lagarde, C. (annual). ECB strategy review (2021) and Sintra Forum speeches.
Bailey, A. (2024). Mansion House speech on inflation and the BoE’s framework review.
Brainard, L. (2024–25, then incoming Fed Chair). NEC and Fed speeches on inflation, financial stability, neutral rate.

Beyond GDP and inequality

Stiglitz, J., Sen, A., Fitoussi, J.P. (2009). Report by the Commission on the Measurement of Economic Performance and Social Progress.
Piketty, T. (2014). Capital in the Twenty-First Century. Harvard Belknap.
Piketty, T., Saez, E., Zucman, G. (2018). “Distributional National Accounts”. QJE 133(2).
Auten, G. and Splinter, D. (2024). “Income Inequality in the United States: Using Tax Data to Measure Long-term Trends”. JPE — the recent revisionist counterpoint.

AI and productivity

Brynjolfsson, E. and McAfee, A. (2014). The Second Machine Age.
Brynjolfsson, E., Li, D., Raymond, L. (2023). “Generative AI at Work”. NBER WP 31161; QJE 2025.
Acemoglu, D. (2024). “The Simple Macroeconomics of AI”. NBER WP 32487.
Acemoglu, D. and Johnson, S. (2023). Power and Progress. PublicAffairs.
Goldman Sachs (Hatzius et al., 2023, 2024 updates). “The Potentially Large Effects of Artificial Intelligence on Economic Growth”.

Textbooks for orientation

Blanchard, O. (2024). Macroeconomics, 9th ed. Pearson. The standard intermediate textbook.
Mankiw, N.G. (2022). Macroeconomics, 11th ed. Worth. The other standard intermediate.
Romer, D. (2019). Advanced Macroeconomics, 5th ed. McGraw-Hill. The standard PhD-prelim text.
Ljungqvist, L. and Sargent, T. (2018). Recursive Macroeconomic Theory, 4th ed. MIT.
Galí, J. (2015). Monetary Policy, Inflation, and the Business Cycle, 2nd ed. Princeton.
Acemoglu, D. (2009). Introduction to Modern Economic Growth. Princeton.
Carlin, W. and Soskice, D. (2014). Macroeconomics: Institutions, Instability, and the Financial System. Oxford — the IS-MP-PC three-equation pedagogical model.
Heer, B. and Maussner, A. (2024). Dynamic General Equilibrium Modeling, 3rd ed. Springer — computational reference.

End of macroeconomics-foundations reference.

Compendium

Explorer

Macroeconomics Foundations — Economics Reference