Demographics Are Driving Inflation, Not Central Banks
Summary
- Economies are complex adaptive systems, characterized by non-linear relationships which change over time. This makes assessing the impact of monetary and fiscal policy difficult.
- Demographics, both population growth and age distribution, impact goods and services and capital markets in a relatively predictable fashion.
- Inflation indices are an aggregate measure that hide as much as they reveal. Inflation may be the result of strong demand / supply constraints generally or for individual products.
- Interest rates are an opportunity cost of holding capital. Lower interest rates generally increase asset prices, but this is dependent on the information content of changes to interest rates.
Inflation has received a lot of focus from investors in recent months as economies globally begin to emerge from the pandemic. There is uncertainty as to whether the current high levels of inflation are the result of base effects and supply chain issues, and hence transitory, or whether they will prove to be more persistent. Much of this conversation seems to be shaded by a belief that central banks are artificially depressing interest rates and recklessly printing money and that inflation or even hyperinflation will be the inevitable result. Looking at the drivers of interest rates and inflation, and digging into aggregates provides insights into why inflation and interest rates have been low for many years. Until the drivers of low interest rates and inflation disappear, persistent inflation will remain unlikely.
Economies are complex adaptive systems where relationships between variables are often dynamic and non-linear, which can make assessing the relationship between variables difficult. Many variables also exhibit trends over long periods which can produce artificially strong correlation, even if this correlation is not meaningful. This is exacerbated by the fact that there are hundreds of variables which affect the economy, many of which are difficult to observe, and there is a limited amount of historical data available to analyze. For these reasons I have not attempted to model any of the relationships discussed and plots are shown only to try and support the discussion. These plots are likely to suffer from the issues discussed above, but when viewed as a mosaic hopefully paint a picture of the current economic conditions and the main factors driving it.
Inflation
While inflation is generally reported as an index, inflation indices are aggregates of a large number of items with weighting based on perceived importance. This is generally not an issue, but during periods of rapid change this can cause indices to be misleading. For example, used car prices have increased rapidly in recent months and have had a large impact on headline inflation. Supply constraints caused by the pandemic are largely responsible for the current levels of inflation and as these constraints are removed inflation is likely to moderate.
Figure 1: US Inflation by Category 1997-2017
(source: marketwatch)
CPI and PCE are two closely followed measures of consumer inflation. There are a number of differences between the indexes, including the formula, weights and scope. The Fed uses PCE as their main measure of inflation making it particularly important for assessing monetary policy. CPI and PCE show similar trends but PCE is less volatile and generally lower than CPI. The University of Michigan consumer inflation expectations index also follows a similar trend to CPI but was generally lower until the mid-90s and has been significantly higher since 2008. This could indicate that consumer expectations are influenced by the Fed's target rate.
Figure 2: US Inflation and Inflation Expectations
(source: Created by author using data from The Federal Reserve)
Another potentially useful measure of inflation breaks CPI into components that are sticky and flexible. Items which are more difficult to reprice and hence are more likely to exhibit price persistence are grouped into the sticky index. The flexible price index appears to be more responsive to the economic environment (slack) while the sticky price index appears to be more forward looking. Forecasts of CPI that are based on the sticky price index tend to be more accurate than forecasts based on CPI. Prior to the COVID pandemic, sticky inflation was building modestly while flexible inflation remained weak. This could indicate that deflationary pressures have moderated in recent years. Sticky inflation has not responded significantly to the pandemic, supporting the belief that inflation is transitory.
Figure 3: US Sticky and Flexible Price Inflation
(source: Created by author using data from The Federal Reserve)
Demographics
Demographics are an important determinant of inflation and interest rates for a number of reasons. A rapidly growing population creates inflationary pressure through increasing demand for goods and services. The age distribution of the population also impacts the economy as the income and expenditure of individuals varies significantly with age. As a result an aging population is more likely to have stagnant demand for goods and services, a higher level of inequality and an excess of savings.
Population Growth
A growing population creates inflationary pressure through increasing demand. Depending on whether growth comes from births or immigration, there may be a substantial lag between an increase in growth and an increase in inflation though. This is because consumption by babies and children is relatively low and only becomes significant as they age. Population growth in most developed countries has been relatively weak in recent decades, easing the inflationary pressure created by the large baby boomer generation.
Figure 4: US Population Growth and Inflation
(source: Created by author using data from The Federal Reserve)
Consumer Saving Lifecycle
A country’s population is not homogeneous, meaning that the age distribution of the population influences the economy in addition to the size of the population. This is because the income and expenditure of individuals varies significantly over the course of their life. Generally, people will have little income until they have finished their education and entered the workforce. Income will then typically rise through the course of an individual’s career, until they retire. Large expenditures that vary with age include higher education, purchasing a home and raising children. Individual’s may begin reducing consumption later in life to help save for retirement and then live within their means during retirement. This is obviously only an illustrative discussion and actual income and expenditure patterns will vary significantly from person to person. Research has shown this effect in practice and as expected demonstrated that the lifecycle effect is muted for low-income individuals. As a result, it would not be unreasonable to assume that life cycle effects are more prominent in high income countries.
Figure 5: Consumption Over the Life Cycle
(source: University of Pennsylvania)
Rising life expectancy could be expected to increase the savings rate while people are working, as they would need to fund a longer period of retirement. Most of the gain in life expectancy has come from reducing premature deaths rather than extending maximum lifespan though, meaning the impact of this is likely to be fairly small.
Figure 6: The Consumer Saving Lifecycle
(source: Created by author)
Inequality and Age Distribution
Due to the fact that consumption and income change over the course of a person’s life, wealth also varies significantly with age. Individuals typically take on debt early in life to finance large purchases like higher education and a home. As their income increases later in life, they payback this debt and begin saving for retirement. Depending on the amount of wealth accrued during a person’s career, wealth could continue to increase, remain stable, or decline during retirement.
Figure 7: Wealth Compounds Over Time
(source: Created by author)
As income and wealth can increase significantly with age in developed countries, it is possible for a country with an aging population to have rising inequality. This rise in inequality is due to age related differences rather than inequality between comparable individuals though and as such is separate to the issue of inequality between the extremes of the income distribution. Global data indicates that inequality is actually lower in countries with older populations, but this could be due to redistributive policies.
Figure 8: Demographics and Inequality Globally
(source: Created by author using data from The World Bank)
Looking at the variation of inequality within countries over time shows that inequality generally remains flat or increases as the population ages, but again this is likely to be influenced by a range of factors.
Figure 9: Demographics and Inequality at the Country Level
(source: Created by author using data from The World Bank)
The recent rise in inequality in the US can largely be modeled using a simple consumer saving lifecycle model that shows rising inequality as the US population has aged. Again, this is separate to inequality due to differences at the extremes of the income distribution and is not trying to attribute inequality purely to demographics, rather simply pointing to the fact that it could be a contributing factor.
Figure 10: Income Inequality in the US and Changes Modelled on Demographics
(source: Created by author using data from The World Bank)
Inflation and Age Distribution
The consumer saving lifecycle shows that aggregate demand is dependent on the age distribution of the population in addition to the size of the population. It is therefore reasonable to expect that as a country’s population ages, aggregate demand will stagnate resulting in weak inflation. Global data shows that inflation generally declines as a country’s median age increases, particularly for nations where the median age is above 40.
Figure 11: Demographics and Inflation Rate Globally
(source: Created by author using data from The World Bank)
Research has shown that the dependent population (young and old) is associated with higher inflation, with the very old (80+ years old) the only exception. This is because working age cohorts are net savers, which drives down inflation, while dependents are net spenders, which drives up inflation. The impending retirement of baby boomers and the increased savings associated with this shift could be responsible for the current low inflation environment in developed economies. If this is the case, recent deflationary pressure from demographics could become inflationary as the baby boomer generation retires, with a potential impact of up to 3-4% by the end of 2030. This was estimated using data from a number of countries between 1870 and 2016. The model matches this data quite well up until the early 2000s when predicted inflation diverges significantly higher than realized inflation. The model also had a weaker fit to data from Greece, Portugal and Japan indicating that the model has issues estimating inflation for older populations. This could indicate that the inflationary pressure of the 50-75 year old age group has been overestimated.
Figure 12: Estimated Inflationary / Deflationary Pressure by Age
(source: sciencedirect)
Interest rates and savings are generally expected to be negatively correlated, but an increase in the dependency ratio leads to both a decline in interest rates and in the saving rate. This is because labor becomes scarcer than capital, even though capital is made scarce by lower savings.
Figure 13: Changing Demographics in the US
(source: Created by author using data from populationpyramid and The Federal Reserve)
Changes in the behavior of individuals over time are likely to impact this analysis though. For example, the labor force participation rate of over 65s had been increasing in recent years, up until the COVID pandemic. All else being equal, a higher participation rate makes this age group more likely to be net savers.
Figure 14: Labor Force Participation Rate of Over 65s in the US
(source: Created by author using data from The Federal Reserve)
Changing demographics is also likely to result in product specific inflationary pressures due to age related demand. For example:
- The very young and the very old are more likely to require healthcare
- Young adults are more likely to undertake higher education
- People in their early 30s are more likely to be first-time home buyers
As the size of these age groups changes it is likely to create inflationary / deflationary pressure for specific products and services.
Figure 15: Excess Healthcare Inflation Rate in the US
(source: Created by author using data from populationpyramid and healthsystemtracker)
Figure 16: Annual Tuition Cost in the US
(source: Created by author using data from populationpyramid and nces)
Figure 17: Home Prices in the US
(source: Created by author using data from populationpyramid and The Federal Reserve)
Conclusion
In recent years a large group of investors has promoted gold, and increasingly Bitcoin (BTC-USD), as a hedge against an imminent economic meltdown caused by reckless monetary policy. If the starting point in analyzing inflation is that hyperinflation is going to occur and then the goal is to find an analogous situation from a developed economy, the Weimar Republic could be acceptable. But if investors look more closely at the current economic conditions, they will find that Japan’s experience in recent decades is more illustrative. Monetary policy is largely responding to forces created by demographic factors and until these forces dissipate or are counter acted, the current environment will likely persist (low inflation, low interest rates, weak economic growth and high asset prices).
This article was written by
Disclosure: I/we have a beneficial long position in the shares of BTC-USD either through stock ownership, options, or other derivatives. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.