Hong Kong Monetary Authority at the International Financial Centre Building, Hong Kong

# Unit 5 Macroeconomic policy: Inflation and unemployment

How governments can moderate costly fluctuations in employment and inflation

### Before you start

This unit builds on the concepts and examples introduced in Units 1, 2, 3, and 4. In particular, you will need to be familiar with the WS–PS model of the supply side of the macroeconomy (Sections 1.5 to 1.7); the multiplier model (Sections 3.6 to 3.8); the Phillips curve (Sections 4.4 to 4.6); and the business cycle model (Sections 4.7 to 4.8).

## 5.1 ‘It’s the economy, stupid’: Government popularity, inflation, and unemployment

The return of high inflation in 2022 after three decades was a shock to families and business managers who had not experienced having to plan their budgets when prices are changing so frequently.

### Inflation and electoral success

best-fit line, line of best fit, best-fit curve
In a scatterplot of two variables x (on the horizontal axis) and y (vertical axis), we can summarize the relationship between y and x by estimating a line of best fit to the set of points. The best-fit line uses the data to predict the average value of y for each value of x. The line slopes upward if x and y are positively correlated, and downward if they are negatively correlated. A best-fit straight line is called a linear regression line; it is also possible to estimate a best-fit curve.
correlation
A statistical association observed between two variables in a sample of data. If high values of one variable (for example, people’s earnings) commonly occur along with high values of another variable (for example, years of education) the variables are positively correlated. When high values of one variable (for example, air pollution) are associated with low values of the other variable (for example, life expectancy) there is a negative correlation. If variables are correlated, it doesn’t mean that there is a causal relationship between them: air pollution may not have caused the lower life expectancy we observed. See also: causality.

Discontent and hardship surrounding inflation can also have an effect on the fortunes of the party in government when an election comes. Figure 5.1 shows the relationship between the margin of victory or defeat for the ruling party in US presidential elections, and the inflation rate, with a line of best fit: the ruling party does better when inflation is low. The two years in which deflation occurred have been omitted from the chart, but if these are included, the negative correlation between the two variables is stronger.

Figure 5.1 Inflation and presidential election victory in the US (1912–2020).

Note: Inflation before 1950: Michael Bordo, Barry Eichengreen, Daniela Klingebiel, and Maria Soledad Martinez-Peria. 2001. ‘Is the Crisis Problem Growing More Severe?’. Economic Policy 16 (32): pp. 52–82; CPI after 1950: Federal Reserve Bank of St. Louis. 2021. FRED; Electoral results: US National Archives. 2021. ‘1789–2021 Presidential Elections’. US Electoral College.

Beyond the US, a study by political scientists has found a relationship between unexpected inflation and electoral outcomes using data from 19 countries.1 More recent studies have highlighted how this relationship depends on countries’ political systems,2 the strategies used by parties and candidates,3 and what conditions are like in other countries.4 Unexpected increases in the cost of living can make governments, as well as households and firms, uncertain about their future prospects.

### Fluctuations in output and unemployment, and the size of government

Policymakers also worry about unemployment. In August 1960, three months before he was elected US president, the 43-year-old Senator John F. Kennedy found time to spend the day cruising Nantucket Sound on his boat, the Marlin. His crew for the day included Seymour Harris, a Harvard economist, and Paul Samuelson, an economist at MIT and later a Nobel laureate. They had not been recruited for their nautical skills—the future president wanted to learn ‘the new economics’, which John Maynard Keynes had formulated in response to the Great Depression.

We explain more about Keynes in Section 5.8.

When Kennedy was a teenager during the Great Depression of the early 1930s, the US and many other countries experienced a drastic fall in output (as shown for the US in Figure 5.2) and massive unemployment that persisted for more than ten years until the Second World War. Kennedy had a lot to learn from his crewmates and Harris later gave him private economics lessons, shuttling by air between Boston, where he worked, and Washington DC.

In 1948, Samuelson had written Economics, the first major textbook to teach these new economic ideas. Harris promoted the same economic ideas in a book called Saving American Capitalism. At that time, it seemed that capitalism needed saving: the centrally planned economies of the Soviet Union and its allies, a model promoted as the alternative to capitalism, had entirely avoided the Great Depression.

Kennedy needed economics to understand policies that could promote economic growth, reduce unemployment, and also avoid economic instability and inflation.

Unit 3 showed that fluctuations in output and employment are a characteristic of capitalist economies. Figure 5.2 shows the annual growth of real GDP in the US economy since 1870.

Figure 5.2 Fluctuations in output and the size of government (as measured by government tax revenue as a percent of national income) in the US (1870–2022).

The Maddison Project. 2020. 2020 Version; US Bureau of Economic Analysis. 2023. GDP & Personal Income; FRED; Wallis, John Joseph. 2000. ‘American Government Finance in the Long Run: 1790 to 1990’. Journal of Economic Perspectives 14 (1) (February): pp. 61–82.

Fluctuations in output

Fluctuations in output were as large as 10–15% of GDP before the 1940s, but became smaller after the end of the Second World War.

Size of the government

Over the same period, the size of the government (as measured by government tax revenue as a share of national income) increased from less than 10% in the late 1800s to more than 30% in the 2020s.

What Harris and Samuelson taught Kennedy was strongly influenced by the key feature displayed in Figure 5.2: the contrast between the volatility of the economy before the Second World War, and the steadier growth and absence of deep recessions afterwards.

In the early part of the chart, the dominant role of agriculture in the economy—a sector affected by weather events and global commodity prices—was one cause of volatility in the economy. But its share of employment fell from 50% in the 1870s to 20% by the late 1930s, yet there was no sign of the economy becoming more stable. Indeed some of the largest fluctuations were during the Great Depression in the 1930s. The big reduction in the magnitude of fluctuations did not occur until after the end of the Second World War.

Figure 5.2 shows another important development: the increasing role of the government in the economy. The red line shows how economists measure the size of the government in the economy: total government tax revenue as a share of GDP.

Note that the ‘size of government’ measures the tax revenue that covers all government expenditure, including transfers. As a result the share of GDP shown is distinctly higher than the share of G in GDP, as shown in Figure 3.5 in Unit 3, which does not include transfer payments.

causal, causality, causation
We can say that a relationship between two variables is causal if we can establish that a change in one variable produces a change in the other. While a correlation is simply an assessment that two things have moved together, causation implies a mechanism accounting for the association, and is therefore a more restrictive concept. See also: natural experiment, correlation.

The fact that the fluctuations in output dramatically reduced while the size of government expanded does not necessarily mean that increased government spending stabilized the economy. There appears to be a relationship between the size of fluctuations and the size of government, but observing that two variables are correlated does not mean that one caused the other. As explained in the box below, statistical correlation does not necessarily mean causation. But there are good reasons to think that the increase in the red line was part of the reason for the smoothing of the black line. In this unit, we shall examine some of these mechanisms.

Macroeconomic policymakers typically aim to limit the impact on wellbeing of the shocks that regularly hit the economy. Without stabilization, the economy can experience persistent high unemployment (as in the Great Depression), a wage–price spiral (as in the 1970s), and stagflation (the 1980s). These episodes are discussed in Unit 9. In this unit, we take the policymaker’s viewpoint and examine macroeconomic policies used to mitigate the effects of shocks.

### Interpreting data: Correlation may not be causation

Can we draw the conclusion from the data in Figure 5.1 that high inflation directly causes governments to lose elections? An obvious alternative explanation might be that high inflation is a symptom of a wider set of problems in the economy (for example, negative supply shocks that have lowered real wages), and that it is these that cause voters to lose confidence in the government.

Similarly, just because the size of the government has risen, does this mean that this has caused the reduction in the volatility of output recorded in Figure 5.2? One line of reasoning, which we shall explore in this unit, is that a larger government can use its tax revenue to provide transfer payments to households, which help them to smooth their consumption. By providing unemployment benefits, for example, the government may reduce the average marginal propensity to consume and therefore, the multiplier. This would reduce the extent to which shocks are transmitted via the multiplier process to employment, income, and consumption. The result would be a less volatile economy.

The marginal propensity to consume and how it is affected by credit constraints is discussed in Section 3.6.

reverse causality
If we are looking for evidence that one variable (x) causes another (y) and find that the variables are correlated, the explanation may be the reverse: that y causes x. For example, if we find that people who attend university earn more, does that mean that university increased their earning ability? Could it be that people with high earning potential are more likely to attend university? See also: correlation.

But maybe we have it the wrong way round, and actually a more stable (and richer) economy allows governments to spend more on health, education, transport, and so on. Economists call this reverse causality.

We always need to be very careful in equating correlation with causation.

Spurious Correlations website and this BBC News article show how dangerous it is to draw a conclusion from correlation.

natural experiment
An empirical study that exploits a difference in the conditions affecting two populations (or two economies), that has occurred for external reasons: for example, differences in laws, policies, or weather. Comparing outcomes for the two populations gives us useful information about the effect of the conditions, provided that the difference in conditions was caused by a random event. But it would not help, for example, in the case of a difference in policy that occurred as a response to something else that might affect the outcome.

To establish a causal relationship between variables, economists may be able to use natural experiments—like the one comparing the growth rates of East and West Germany described in Section 1.10 of the microeconomics volume—or other statistical techniques. Controlled experiments, described in Section 4.9 of the microeconomics volume, are rarely possible in macroeconomics.

### Exercise 5.1 Correlation vs causation

Even though two variables are strongly correlated with each other, it is not necessarily the case that one variable’s behaviour is the result of the other (a characteristic known as causation⁠). The two variables could be spuriously correlated. The following example illustrates spurious correlation⁠:

A child’s academic performance may be positively correlated with the number of rooms in their house or house size, but if the child already has space to study, it seems implausible that building an extra room would make a child smarter, or that doing well at school would make your house bigger. It is more plausible that income or wealth, which determines the size of home that a family can afford and the resources available for studying, is the ‘unseen factor’ in this relationship.

Choose two examples of spurious correlation from Tyler Vigen’s Spurious Correlations website. Explain whether you think the observed relationship between the two variables is a coincidence, or whether this correlation could be due to one or more other variables (potential ‘unseen factors’).

1. H. D. Palmer and G. D. Whitten. 1999. ‘The Electoral Impact of Unexpected Inflation and Economic Growth’. British Journal of Political Science 29 (4): pp. 623–639.

2. Åsa Bengtsson. 2004. ‘Economic Voting: The Effect of Political Context, Volatility and Turnout on Voters’ Assignment of Responsibility’. European Journal of Political Research 43: pp. 749–767.

3. Timothy Hellwig. 2012. ‘Constructing Accountability: Party Position Taking and Economic Voting’. Comparative Political Studies 45 (1): pp. 91–118.

4. Selim E. Aytaç. 2018. ‘Relative Economic Performance and the Incumbent Vote: A Reference Point Theory’. The Journal of Politics 80 (1): pp. 16–29.