Macroeconomics — aggregates, growth, and cycles

Intuition Beginner

Microeconomics 52.01.01 studies one market at a time. Macroeconomics steps back and asks how the whole economy behaves: total output, the average level of prices, and the number of people with jobs. The lens widens from a single buyer and seller to an entire nation's accounts.

Three quantities anchor the field. Output — gross domestic product, or GDP — is the total value of goods and services a country produces in a year. Inflation is the rate at which the general price level rises. Unemployment is the share of people who want work but cannot find it.

Two questions organise the subject. Over decades, why do some countries grow rich while others stay poor? Over a few years, why does output rise and fall in booms and recessions? Long-run growth and short-run cycles are the two halves of macroeconomics.

Governments and central banks watch these aggregates because their tools — taxes, spending, interest rates — move the whole economy at once, not just one market.

Visual Beginner

The three core indicators and how they typically move across the business cycle:

Indicator	Expansion	Recession
GDP growth	Above trend	Below trend, or negative
Unemployment	Falling	Rising
Inflation	Often rising	Often easing

The straight trend line is long-run growth; the wiggles around it are the cycle.

Worked example Beginner

The Solow growth model asks how much capital (machines, factories, roads) a country accumulates, and what level it settles at. More saving builds more capital, but each year some capital wears out and must be shared among more workers.

Use these numbers. Saving rate $s=0.2$ means 20% of income is saved. Depreciation $\delta =0.05$ means 5% of the capital stock wears out yearly. Population growth $n=0.02$ and technology growth $g=0.02$ are both 2% per year. Capital share $\alpha =0.3$.

Step 1. Add the three shrinkers: $\delta +n+g=0.05+0.02+0.02=0.09$.

Step 2. Divide saving by that total: $0.2/0.09\approx 2.22$.

Step 3. Raise to the power $1/(1-\alpha )=1/0.7\approx 1.43$. The steady-state capital per effective worker is $k^{\ast }\approx 2.22^{1.43}\approx 3.1$.

At $k^{\ast }$ the economy stops growing in per-worker terms: new investment exactly offsets wear-out and dilution. Output per effective worker settles at $y^{\ast }=(k^{\ast }{)}^{\alpha }\approx 3.1^{0.3}\approx 1.4$. Only faster technology growth ($g$) lifts living standards in the long run.

Check your understanding Beginner

Formal definition Intermediate+

Let $Y_t$ denote real GDP, $P_t$ the price level, and $L_t$ the labour force. The national-income identity is

$$Y_t=C_t+I_t+G_t+N{X}_t,N{X}_t=X_t-M_t,$$

with consumption $C$, investment $I$, government purchases $G$, and net exports $NX$. The inflation rate and the unemployment rate are

$${\pi }_t=\frac{P_t-P_{t-1}}{P_{t-1}},u_t=\frac{{\text{unemployed}}_t}{L_t}.$$

Solow growth model. Output is produced from capital $K_t$ and effective labour $A_t{L}_t$ by a Cobb-Douglas production function $Y_t=K_t^{\alpha }(A_t{L}_t{)}^{1-\alpha }$ with $0<\alpha <1$, where $A_t$ grows at rate $g$ and $L_t$ at rate $n$. Capital accumulates according to ${\dot{K}}_t=s{Y}_t-\delta K_t$, where $s$ is the constant saving rate and $\delta$ the depreciation rate. Writing capital per effective worker as $k_t=K_t/(A_t{L}_t)$, the law of motion collapses to

$${\dot{k}}_t=s{k}_t^{\alpha }-(\delta +n+g)k_t.$$

The aggregate-demand / aggregate-supply (AD-AS) framework pairs a downward-sloping AD curve (from the IS-LM equilibrium) with a short-run aggregate supply (SRAS) that slopes up when some prices are sticky and a vertical long-run aggregate supply (LRAS) at potential output $\bar{Y}$.

The IS-LM model equates a goods-market equilibrium (IS) $Y=C(Y-T)+I(r)+G$ with a money-market equilibrium (LM) $M/P=L(Y,r)$, jointly determining output $Y$ and the real interest rate $r$ for given fiscal policy $(G,T)$ and real money supply $M/P$.

The expectations-augmented Phillips curve relates inflation ${\pi }_t$, expected inflation ${\pi }_t^e$, and the gap between actual unemployment and the natural rate $u_n$:

$${\pi }_t={\pi }_t^e-\beta (u_t-u_n),\beta >0.$$

Counterexamples to common slips

• GDP is not the same as welfare. It excludes non-market activity and leisure, and says nothing about distribution, so GDP growth is a measure of output, not a measure of a good society.
• A higher saving rate does not raise growth forever. In the Solow model a higher $s$ raises the level of the steady state but not its long-run growth rate, which equals $g$.
• The Phillips curve is not a stable menu. Once expectations adjust, the short-run inflation-unemployment tradeoff disappears, leaving the long-run curve vertical at $u_n$.

Economic theory Intermediate+

The central object of long-run macro is the steady state of the Solow model ^{[Solow 1956]}.

Theorem (Solow steady state and convergence). In the Solow model with $f(k)=k^{\alpha }$, $0<\alpha <1$, and $s,\delta ,n,g>0$, there is a unique positive steady state

$$k^{\ast }={\left(\frac{s}{\delta +n+g}\right)}^{1/(1-\alpha )},$$

and from any initial $k_0>0$ the economy converges to $k^{\ast }$.

Argument sketch. The steady state solves $s(k^{\ast }{)}^{\alpha }=(\delta +n+g)k^{\ast }$, giving the displayed formula. The law of motion $\dot{k}=s{k}^{\alpha }-(\delta +n+g)k$ is positive when $k<k^{\ast }$ and negative when $k>k^{\ast }$, because $k^{\alpha }$ is strictly concave. Hence the vector field points toward $k^{\ast }$ everywhere, so the steady state is globally asymptotically stable. Linearising around $k^{\ast }$ gives a local convergence rate of $(1-\alpha )(\delta +n+g)$. At the steady state, per-effective-worker variables are constant and total output grows at rate $n+g$; per-capita output grows at rate $g$. $\square$

Comparative positions on stabilization policy. Macroeconomics is unusual among quantitative fields in that its core policy question — whether governments and central banks should actively manage aggregate demand — is genuinely contested.

The Keynesian position holds that sticky wages and prices make output demand-determined in the short run, so countercyclical fiscal and monetary policy can reduce the welfare cost of recessions ^{[Keynes 1936]}. On this view the 2008–2009 fiscal stimulus and the post-2020 rate cuts were appropriate responses to demand shortfalls.

The classical and monetarist counterposition holds that economies self-correct toward potential output, that the long and variable lags of policy and the re-optimisation emphasised by the Lucas critique make activist intervention destabilising, and that stable rules (such as money-growth or inflation-targeting rules) outperform discretion ^{[Lucas 1976]}. The Volcker disinflation and rule-based inflation targeting are cited in this tradition.

The models underdetermine the choice: the positive claims about price stickiness and policy lags are empirical, and the normative claim that government ought to stabilise the cycle does not follow from any model alone. It rests on value judgements about the relative costs of unemployment and inflation, and on political views about the proper scope of state action.

Bridge. This steady-state result builds toward 52.03.01 (econometrics), where growth and convergence regressions test the prediction against cross-country data, and appears again in 44.08.03 (Markov decision processes), whose Bellman-equation machinery is exactly the intertemporal-optimisation foundation that replaces the exogenous saving rate in modern growth theory. The foundational reason the steady state matters is that it separates long-run growth (driven by technology $g$) from transitional dynamics, and putting these together, the bridge is the mapping from a parameterised law of motion to a stationary long-run outcome — the pattern every macro model in the rest of this unit reuses.

Exercises Intermediate+

Lean formalization Intermediate+

lean_status: none. Macroeconomics is a model-and-evidence discipline; its correctness gate is internal consistency, calibration against national accounts, and empirical fit, not formal proof. The dynamical-systems machinery (fixed points, linearisation, stability analysis) that underlies the Solow convergence argument is the subject of 44.08.03, and the constrained-optimisation foundations (Lagrangian, KKT, Bellman, dynamic programming) live in 44.02.01. Those mathematical layers, not the macroeconomic models built on top of them, are the natural targets for formalisation.

Advanced results Master

The Solow model is the workhorse of long-run macro, but it leaves the growth rate $g$ unexplained — it falls out of the sky as "exogenous" technical progress. Endogenous growth theory (Romer 1990; Aghion and Howitt 1992) endogenises $g$ by modelling research-and-development as a deliberate activity that produces non-rival knowledge. Because knowledge can be used by many firms at once, the production of ideas exhibits increasing returns at the economy level, and the long-run growth rate becomes a function of the incentives to innovate — the size of the market, the rate of time preference, and the strength of intellectual-property protection. The micro-foundations draw on the firm theory of 52.01.01.

Real business cycle (RBC) theory (Kydland and Prescott 1982) carries the micro-foundations programme to the short run. It models the economy as a single representative household that optimises over time under flexible prices and budget constraints, with business cycles driven entirely by technology shocks. In RBC models monetary policy is neutral and cycles are the efficient response of the economy to real disturbances — a radical departure from the Keynesian view of cycles as demand failures requiring a policy remedy.

The New Keynesian response re-derives the Keynesian sticky-price conclusions from micro-foundations: monopolistically competitive firms set prices subject to a Calvo-style adjustment friction, and the resulting three-equation model — a forward-looking IS equation, a New Keynesian Phillips curve, and a monetary-policy rule — reproduces the IS-LM and Phillips-curve relationships as the equilibrium of an optimisation problem ^{[Woodford 2003]}. The dynamic-stochastic-general-equilibrium (DSGE) framework that combines RBC real forces with New Keynesian nominal rigidities is the contemporary synthesis, the workhorse of central-bank forecasting and policy analysis.

Synthesis. Modern macroeconomics is the micro-founded dynamic synthesis of growth and cycles: the foundational reason the field rebuilt itself around representative-agent optimisation in the 1980s and 1990s is that aggregate relationships without choice are not policy-invariant, this is exactly the content of the Lucas critique, and the central insight of the New Keynesian DSGE framework is that sticky-price monopolistic competition plus intertemporal optimisation reproduces the IS-LM and Phillips-curve relationships on micro-foundations. Putting these together, the bridge is that the Solow growth model, the IS-LM short-run model, and the New Keynesian policy model are all instances of one pattern — a dynamic system whose steady state pins down long-run outcomes and whose transitional dynamics describe the cycle — and the pattern generalises to the heterogeneous-agent and financial-friction models that drive current research.

Full proof set Master

Proposition (Speed of convergence in the Solow model). Consider the Solow law of motion $\dot{k}=s{k}^{\alpha }-(\delta +n+g)k$ with $0<\alpha <1$ and $s,\delta ,n,g>0$, and let $k^{\ast }={\left(s/(\delta +n+g)\right)}^{1/(1-\alpha )}$ be its unique positive steady state. Then in a neighbourhood of $k^{\ast }$ the log-deviation ${\hat{k}}_t=\ln (k_t/k^{\ast })$ evolves according to ${\dot{\hat{k}}}_t=-(1-\alpha )(\delta +n+g){\hat{k}}_t+O({\hat{k}}_t^2)$, so the half-life of a deviation from $k^{\ast }$ is $\ln 2/((1-\alpha )(\delta +n+g))$.

Proof. Rewrite the law of motion in growth-rate form,

$$\frac{\dot{k}}{k}=s{k}^{\alpha -1}-(\delta +n+g).$$

Define $\phi (\ln k)=s{e}^{(\alpha -1)\ln k}-(\delta +n+g)$, so that $\dot{k}/k=\phi (\ln k)$. At the steady state $\phi (\ln k^{\ast })=0$. Differentiating,

$${\phi }^{\prime }(\ln k^{\ast })=(\alpha -1)s(k^{\ast }{)}^{\alpha -1}.$$

Using the steady-state identity $s(k^{\ast }{)}^{\alpha -1}=\delta +n+g$, this becomes ${\phi }^{\prime }(\ln k^{\ast })=-(1-\alpha )(\delta +n+g)$, which is strictly negative. Since ${\dot{\hat{k}}}_t=\phi (\ln k_t)=\phi (\ln k^{\ast }+{\hat{k}}_t)$, a first-order Taylor expansion around ${\hat{k}}_t=0$ gives ${\dot{\hat{k}}}_t={\phi }^{\prime }(\ln k^{\ast }){\hat{k}}_t+O({\hat{k}}_t^2)$, which is the displayed linear dynamics. The coefficient $-(1-\alpha )(\delta +n+g)$ is negative, confirming local asymptotic stability, and a deviation decays to half its size in time $\ln 2/((1-\alpha )(\delta +n+g))$. $\square$

For the calibrated values $\alpha =0.3$, $\delta +n+g=0.09$, the convergence rate is $0.7\times 0.09=0.063$ per year and the half-life is about $11$ years — the empirical regularity that motivates conditional convergence: economies converge to their own steady states, but only slowly, so cross-country income differences persist for decades.

Connections Master

• Microeconomics 52.01.01. Every macro relationship is built from micro-decisions: the consumption function aggregates household optimisation, investment aggregates firm optimisation, and the New Keynesian Phillips curve aggregates the price-setting of monopolistically competitive firms. Modern macro is explicitly micro-founded, so the constrained-optimisation and equilibrium tools of microeconomics are the load-bearing machinery of this unit.

• Econometrics 52.03.01. The predictions of macro models — the convergence regression, the Phillips-curve slope, the estimated interest-sensitivity of investment — are tested and quantified with econometric methods. Identification of causal policy effects (the natural-rate hypothesis, the effects of monetary shocks) is the central methodological problem where macro meets econometrics.

• Game theory 52.04.01. Strategic interaction underlies the time-inconsistency of optimal monetary and fiscal plans (Kydland-Prescott 1977), the credibility of central-bank commitments, and the political-economy of stabilisation. Repeated-game analysis formalises how reputation can substitute for binding policy rules.

• Optimisation and dynamic systems 44.02.01 and 44.08.03. The differential-equation and fixed-point machinery of the Solow convergence proof, and the Bellman-equation dynamic programming that underpins modern micro-founded models, are the formal subjects of the optimisation and dynamic-programming chapters.

Historical & philosophical context Master

Macroeconomics as a distinct field begins with Keynes's The General Theory of Employment, Interest and Money (1936), written against the mass unemployment of the Great Depression. Keynes argued that aggregate demand could settle at a level too low to sustain full employment, and that fiscal policy could close the gap — the principle of effective demand ^{[Keynes 1936]}. The "Keynesian revolution" built the IS-LM apparatus (Hicks 1937), the consumption function, and the Phillips-curve tradeoff (Phillips 1958) into the postwar neoclassical synthesis, which married Keynesian short-run dynamics with classical long-run growth.

Long-run growth was formalised by Solow (1956) and Swan (1956). The neoclassical growth model pinned the steady state to saving, depreciation, population, and technology, and implied conditional convergence across economies ^{[Solow 1956]}. Solow's 1957 follow-up found that the bulk of historical US growth was accounted for not by capital deepening but by the unexplained "residual" of technical progress — founding modern growth accounting and motivating the endogenous-growth programme that followed forty years later.

The 1970s stagflation — high inflation and high unemployment together — fractured the synthesis. Lucas (1976) argued that aggregate relationships are not invariant to policy change because rational agents re-optimise when the policy rule changes: the Lucas critique ^{[Lucas 1976]}. Kydland and Prescott (1982) built the real-business-cycle model on flexible prices and technology shocks; the New Keynesian response (Woodford 2003) re-derived sticky-price Keynesian results from explicit micro-foundations ^{[Woodford 2003]}. The DSGE framework that emerged from this exchange is the contemporary working synthesis of growth and cycle theory.

The standing philosophical tension is whether macroeconomics is a positive science of aggregate regularities or a normative programme for managing capitalism. The models themselves are positive; the policy advice drawn from them is irreducibly normative, and rests on judgements about the costs of unemployment, the legitimacy of state intervention, and the trade-off between stability and freedom that no empirical estimate can settle.

Bibliography Master

@book{Mankiw2021,
  author = {Mankiw, N. Gregory},
  title = {Macroeconomics},
  edition = {11},
  publisher = {Worth Publishers},
  year = {2021},
}

@book{Blanchard2017,
  author = {Blanchard, Olivier},
  title = {Macroeconomics},
  edition = {7},
  publisher = {Pearson},
  year = {2017},
}

@book{Romer2018,
  author = {Romer, David},
  title = {Advanced Macroeconomics},
  edition = {5},
  publisher = {McGraw-Hill},
  year = {2018},
}

@article{Solow1956,
  author = {Solow, Robert M.},
  title = {A Contribution to the Theory of Economic Growth},
  journal = {Quarterly Journal of Economics},
  volume = {70},
  number = {1},
  pages = {65--94},
  year = {1956},
}

@book{Keynes1936,
  author = {Keynes, John Maynard},
  title = {The General Theory of Employment, Interest and Money},
  publisher = {Macmillan},
  year = {1936},
}

@incollection{Lucas1976,
  author = {Lucas, Robert E.},
  title = {Econometric Policy Evaluation: A Critique},
  booktitle = {The Phillips Curve and Labor Markets},
  editor = {Bruner, K. and Meltzer, A.},
  series = {Carnegie-Rochester Conference Series},
  publisher = {North-Holland},
  pages = {19--46},
  year = {1976},
}

@book{Woodford2003,
  author = {Woodford, Michael},
  title = {Interest and Prices: Foundations of a Theory of Monetary Policy},
  publisher = {Princeton University Press},
  year = {2003},
}

@article{KydlandPrescott1982,
  author = {Kydland, Finn E. and Prescott, Edward C.},
  title = {Time to Build and Aggregate Fluctuations},
  journal = {Econometrica},
  volume = {50},
  number = {6},
  pages = {1345--1370},
  year = {1982},
}