前四章研究了个人和企业在具体市场中如何做出决策。现在我们转换视角。宏观经济学研究整体经济——商品和服务的总产出、总体价格水平、失业率以及构成商业周期的扩张与收缩模式。
在分析这些现象之前，我们必须先对其进行测量。本章介绍量化总体经济活动的国民核算框架。数字本身不是重点——重点在于它们揭示了经济如何运行，以及它们掩盖了什么。
这个定义中有四个关键词值得深究：
经济循环流量保证了GDP可以通过三种等价方式来衡量：
1. 支出法：加总所有最终商品和服务的支出。
| 组成部分 | 包含内容 | 典型份额 |
|---|---|---|
| $C$ — 消费 | 家庭在商品和服务上的支出 | ~60–70% |
| $I$ — 投资 | 企业固定投资、住宅投资、存货变动 | ~15–20% |
| $G$ — 政府支出 | 政府购买商品和服务（不含转移支付） | ~15–20% |
| $NX$ — 净出口 | 出口减去进口 | 不定（可为负） |
2. 收入法：加总生产过程中获得的所有收入。
花在最终商品上的每一美元都会成为某人的收入——工人的工资、房东的租金、贷方的利息、所有者的利润。
3. 生产（增加值）法：汇总生产各环节的增加值。
如果农民种植小麦（\$1），磨坊主磨面粉（\$1），面包师卖面包（\$1），增加值为：\$1 + (\$1 − \$1) + (\$1 − \$1) = \$1 + \$1 + \$1 = \$1 = 最终产品价格。
三种方法得出相同的GDP——这是一个会计恒等式，而非理论。
悬停在箭头上查看每个流动的描述。四个部门——家庭、企业、政府和外国部门——通过产品市场和要素市场连接。
图7.1.循环流量图。每一美元的支出（C、I、G、NX）都成为收入（工资、租金、利息、利润）。政府征收税款并进行转移支付。外国部门增加出口并减去进口。
几个边界案例有助于澄清GDP的概念：
名义GDP可能因为经济生产了更多东西或价格上涨而增长。为了衡量实际产出增长，我们需要将两者分离。
What this says: The GDP deflator tells you how much of nominal GDP growth is just price increases rather than real production. If nominal GDP doubled but real GDP stayed the same, the deflator doubled — all that "growth" was inflation.
Why it matters: It lets you strip away inflation to see whether an economy is actually producing more goods and services, or just charging more for the same output.
See Full Mode for the derivation.通货膨胀率是价格指数的百分比变化：
| 特征 | CPI | GDP平减指数 |
|---|---|---|
| 商品篮子 | 固定的（消费品） | 所有国内生产的商品 |
| 进口商品 | 包含（消费者购买这些商品） | 不包含（非国内生产） |
| 新商品 | 纳入较慢 | 自动纳入 |
| 替代偏差 | 有（固定篮子高估通胀） | 无（篮子会调整） |
一个经济体只生产两种商品：苹果和电脑。
| 第1年（基年） | 第2年 | |||
|---|---|---|---|---|
| 价格 | 数量 | 价格 | 数量 | |
| 苹果 | \$1 | 100 | \$1.50 | 80 |
| 电脑 | \$100 | 10 | \$100 | 15 |
名义GDP：第1年：\$1(100) + \$100(10) = \$1,100。第2年：\$1.50(80) + \$100(15) = \$1,120。
实际GDP（第1年价格）：第2年：\$1(80) + \$100(15) = \$1,580。
GDP平减指数（第2年）： \$1,120 / \$1,580 × 100 = 80.7。价格水平下降了，因为电脑价格下降的幅度超过了苹果价格上涨的幅度。
You now have the tools to measure national output and compare living standards across countries. Here's what those numbers reveal about the biggest question in economics — and what they can't tell you yet.
GDP per capita is the standard metric for comparing living standards. At purchasing power parity, the United States produces roughly \$80,000 per person per year. India produces about \$9,000. The Democratic Republic of Congo produces around \$600. That's a ratio of more than 130 to 1 between the richest and poorest countries. The gap is enormous, persistent, and — once you adjust for PPP — not an artifact of exchange rate fluctuations. Real GDP per capita, imperfect as it is, captures something fundamentally real about the difference in material living standards across nations.
GDP misses enormous swaths of economic life. In many developing countries, 30–60% of economic activity occurs in the informal sector — subsistence farming, street markets, home production — none of which appears in official statistics. GDP ignores the distribution of income within countries: a nation's GDP per capita can rise while most of its citizens get poorer, if gains concentrate at the top. It excludes environmental degradation, unpaid care work, and leisure. Alternative measures like the Human Development Index, which adds health and education, tell different stories — Cuba ranks far above its GDP bracket on HDI, while oil-rich Equatorial Guinea ranks far below. The gap is real, but GDP alone overstates the gap in some dimensions and understates it in others.
The development economics mainstream uses GDP per capita as a starting point while acknowledging its limitations. The World Bank and IMF supplement GDP with health metrics (life expectancy, infant mortality), education metrics (years of schooling, literacy), and inequality measures (Gini coefficient). PPP adjustments partially address price-level differences. But the profession treats GDP as indispensable because no alternative measure is simultaneously comprehensive, comparable across countries, and available at high frequency. The gap is real even after every reasonable adjustment.
GDP per capita is an imperfect but irreplaceable starting point. The 50-to-1 income gap between rich and poor countries is real, robust to measurement adjustments, and one of the most important facts in economics. Whether you supplement GDP with HDI, multidimensional poverty indices, or satellite nightlight data, the basic picture holds: some countries produce vastly more per person than others, and this translates into enormous differences in health, education, and life expectancy. The question is not whether the gap exists — it's why.
Measurement tells you the gap exists. It does not tell you why. Is it capital accumulation? Ideas and technology? Institutions? Geography? Culture? You need causal models, not accounting identities. Come back at Chapter 9 (§9.4), where the Solow model offers the first causal story — and immediately reveals its own inadequacy. Then Chapter 13 adds endogenous growth, Chapter 18 adds institutions, and Chapter 20 confronts the empirical frontier. The answer, you'll find, is genuinely contested.
GDP per capita correlates with life satisfaction up to about \$75,000 — then the relationship flattens. Does that mean GDP is the wrong target, or that rich countries have already won?
入门其中$U$是失业人数，$E$是就业人数，$L = U + E$是劳动力。
失业率每高于自然失业率一个百分点，就对应约2%的产出损失。系数（2）是一个经验估计值，因国家和时期而异。
What this says: When unemployment rises 1 percentage point above its "normal" level, the economy loses roughly 2% of its potential output. The relationship is roughly 2-to-1: each point of excess unemployment costs about two points of GDP.
Why it matters: It puts a dollar figure on joblessness. A recession that pushes unemployment 3 points above normal wastes about 6% of what the economy could produce — trillions of dollars in a large economy.
See Full Mode for the derivation.一个经济体的$u_n = 5\%$，潜在GDP为$Y^* = \\$10\text{B}$，实际失业率为$u = 7\%$。
产出缺口：$\frac{Y - Y^*}{Y^*} \approx -2(0.07 - 0.05) = -4\%$
实际GDP：$Y \approx 0.96 \times \\$10\text{B} = \\$1.6\text{B}$
经济产出比潜在水平低4亿美元——这是2个百分点周期性失业的代价。
一个国家报告了以下数据（单位：十亿）：家庭消费 = \$100，企业投资 = \$150，政府购买 = \$100，出口 = \$100，进口 = \$120。
支出法： $Y = C + I + G + NX = 600 + 150 + 200 + (100 - 120) = \\$130\text{B}$
各组成部分份额： C = 64.5%，I = 16.1%，G = 21.5%，NX = −2.2%。
收入法将得出相同的\$130B，通过加总工资（\$150B）、租金（\$10B）、利息（\$10B）、利润（\$100B）、折旧（\$10B）和间接税（\$10B）。
生产法汇总各行业的增加值——农业（\$10B）、制造业（\$150B）、服务业（\$130B）= \$130B。
根据循环流量恒等式，三种方法得出相同的GDP。
| 阶段 | 描述 |
|---|---|
| 扩张 | 实际GDP上升；就业增长；生产增加 |
| 峰值 | 衰退前的最高点 |
| 收缩（衰退） | 实际GDP下降；就业减少；生产下降 |
| 谷底 | 复苏前的最低点 |
图7.2.商业周期描述了GDP围绕其长期增长趋势的短期波动。将鼠标悬停在GDP线上，查看每个时点的阶段。
| 分类 | 含义 | 示例 |
|---|---|---|
| 顺周期 | 扩张时上升，衰退时下降 | GDP、消费、投资、就业 |
| 逆周期 | 扩张时下降，衰退时上升 | 失业率 |
| 非周期性 | 无系统性模式 | 政府支出（取决于政策） |
关键规律性事实：
| 变量 | $\sigma_x / \sigma_Y$ | 解释 |
|---|---|---|
| GDP（$Y$） | 1.00 | 参照基准 |
| 消费（$C$） | 0.5 | 波动性为GDP的一半——消费平滑 |
| 投资（$I$） | 3.0 | 波动性为GDP的三倍——放大器 |
| 工作小时数 | 0.8 | 波动性与产出接近 |
| 实际工资 | 0.4 | 相对平滑 |
You now know what recessions look like — output falls, investment collapses, unemployment rises. But knowing the symptoms is not the same as knowing the disease. What actually causes these episodes?
The stylized facts you just learned are remarkably stable: recessions involve GDP falling from peak to trough, investment dropping 3–4 times more than output, consumption declining modestly (households smooth), and unemployment rising with a lag. These regularities have held from the Great Depression through the 2008 financial crisis and the 2020 pandemic recession. They are empirical observations — descriptions of what happens — not explanations of why. The business cycle is a real phenomenon, not a statistical artifact. Something causes output to deviate from trend, and the pattern is too regular to be random noise.
Even the measurement is contested. GDP captures market production, not welfare. A recession that shifts activity into home production, the informal sector, or leisure may be less severe than the GDP numbers suggest. The NBER's business cycle dating is retrospective and involves judgment — recessions are declared months after they begin. More fundamentally, is the "business cycle" even a coherent object? Some economists argue that each recession has a unique cause — oil shocks in the 1970s, monetary tightening in 1981, a financial crisis in 2008, a pandemic in 2020 — and that searching for a unified theory of recessions is like searching for a unified theory of car accidents.
The stylized facts have been stable since Burns and Mitchell documented them in 1946. What has changed is the interpretation. The Keynesian tradition sees cycles as deviations from a stable trend — the economy falls below its potential because of insufficient demand. The Real Business Cycle tradition, starting in the 1980s, sees cycles as movements of the trend itself — productivity shocks shift what the economy can optimally produce. This is not a minor distinction. It determines whether recessions are market failures that demand policy intervention or efficient responses that policy should leave alone.
At this stage, you have the facts but not the theory. That's the right starting point. The regularities — consumption smoothing, investment volatility, countercyclical unemployment — are agreed upon across all schools of thought. What the schools disagree about is interpretation: are recessions demand failures, supply adjustments, financial panics, or some combination? You need models to evaluate these competing claims, and you don't have them yet. Know the facts cold before hearing the theories.
Three competing explanations await. In Chapter 8 (§8.1, §8.8), the Keynesian cross and AD-AS model say recessions are caused by insufficient aggregate demand — a fall in confidence or spending that the sticky-price economy can't quickly absorb. In Chapter 14, Real Business Cycle theory says recessions are optimal responses to negative technology shocks — the economy is always in equilibrium. In Chapter 15, the New Keynesian synthesis tries to combine both stories into a single framework. The profession still does not fully agree on which story is right.
The yield curve inverted, leading indicators are mixed, and forecasters disagree. You now know what a recession looks like in the data — can you spot one before it arrives?
中级支出恒等式$Y = C + I + G + NX$可以重新排列，揭示储蓄、投资和贸易之间的基本关系。
私人储蓄： $S_{private} = Y - T - C$
公共储蓄： $S_{public} = T - G$
国民储蓄： $S = S_{private} + S_{public} = Y - C - G$
根据支出恒等式：
这就是储蓄-投资恒等式：国民储蓄与国内投资之差等于净出口。一个储蓄多于投资的国家有贸易顺差；一个投资多于储蓄的国家必须从国外借款，存在贸易逆差。
What this says: Every dollar a country earns but doesn't consume or hand to the government is "saved." That saving either finances domestic investment (building factories, houses) or flows abroad as a trade surplus. If a country invests more than it saves, the difference must come from foreign borrowing — which shows up as a trade deficit.
Why it matters: A trade deficit isn't inherently bad — it can mean a country is attracting investment because it has great opportunities. Conversely, a government budget deficit can crowd out trade by absorbing national saving, creating the "twin deficits" pattern.
See Full Mode for the derivation.调整GDP的各个组成部分，实时观察支出恒等式、净出口、国民储蓄和S−I=NX恒等式的变化。
图7.1.GDP组成部分堆叠柱状图。净出口可能为负，显示在零线以下。右侧柱子分解了国民储蓄和投资，验证$S - I = NX$。
调整两种商品的价格和数量。观察名义GDP、实际GDP、GDP平减指数和通货膨胀率如何响应。注意通货膨胀如何在实际产出下降时仍使名义GDP增长。
图7.3.比较名义GDP（现行价格）和实际GDP（基年价格）。两者之间的差距反映了GDP平减指数所捕捉的总体价格水平变化。
滑动失业率，观察产出缺口和实际GDP的响应。奥肯定律：失业率每高于自然失业率（$u_n$）一个百分点，潜在GDP就损失约2%。
图7.4.潜在GDP与实际GDP。阴影区域代表因周期性失业而损失的产出。当$u = u_n$（5%）时，缺口为零，经济在潜在水平运行。
凯拉尼共和国是一个拥有500万人口的小岛国。在后续宏观经济学章节中，我们将以凯拉尼作为应用理论的实验室。
国民账户（第1年，十亿KD）： C = 7.0，I = 1.5，G = 2.5，X = 2.0，M = 1.0。
GDP = 7.0 + 1.5 + 2.5 + (2.0 − 1.0) = 100亿KD。人均GDP：2,000 KD。
测量挑战：凯拉尼有庞大的非正规经济部门（约占经济活动的30%）。真实GDP可能接近130亿KD。
劳动力市场：劳动年龄人口：350万。劳动力：280万（劳动参与率 = 80%）。失业人数：28万。失业率：$u = 10\%$。
奥肯定律：如果$u_n = 7\%$且$Y^* = 10.5$B KD，产出缺口$\approx -2(0.10 - 0.07) = -6\%$。预测实际GDP：\$1.94 \times 10.5 = 9.87$B KD。测量GDP为10.0B——说明自然失业率估计偏低，或者凯拉尼的奥肯系数有所不同。
Maya的柠檬水摊每天收入\$123.75（第2章）将通过支出法计入GDP——这是她的顾客的消费支出。但如果Maya不申报收入，它就落入非正规经济，被官方统计遗漏——这正是凯拉尼30%非正规经济部门所面临的测量挑战。
| 标签 | 公式 | 描述 |
|---|---|---|
| 等式 7.1 | $Y \equiv C + I + G + NX$ | 支出恒等式 |
| 等式 7.2 | $Y \equiv$ 工资 + 租金 + 利息 + 利润 + ... | 收入恒等式 |
| 等式 7.3 | 增加值 = 收入 − 中间投入 | 生产法 |
| 等式 7.4 | 实际GDP$_t = \sum P_i^{base} \times Q_i^t$ | 按基年价格计算的实际GDP |
| 等式 7.5 | GDP平减指数 = (名义GDP / 实际GDP) × 100 | GDP平减指数 |
| 等式 7.6 | CPI$_t$ = (第t年一篮子商品成本 / 基期一篮子商品成本) × 100 | 消费者价格指数 |
| 等式 7.7 | $\pi_t = (P_t - P_{t-1})/P_{t-1} \times 100$ | 通货膨胀率 |
| 等式 7.8 | $u = U / (U + E)$ | 失业率 |
| 等式 7.9 | $LFPR = L / \text{劳动年龄人口}$ | 劳动参与率 |
| 等式 7.10 | $(Y - Y^*)/Y^* \approx -2(u - u_n)$ | 奥肯定律（水平形式） |
| 等式 7.11 | $\Delta Y/Y \approx 3\% - 2\Delta u$ | 奥肯定律（增长形式） |
| 等式 7.12 | $S = I + NX$ | 储蓄-投资恒等式 |
| 等式 7.13 | $S - I = NX$ | 贸易差额 = 储蓄缺口 |