Baby booms and the housing market: is there a bust on the horizon?

'We find that birth rates today strongly predict house price increases around 30 years later'

In our paper, ‘Baby Booms and Asset Booms: Demographic Change in the Housing Market,’ we revisit a question that has occupied economists for decades: to what extent do fluctuations in housing markets result from changes in population and age structure? Answering this question has been challenging because demographic shifts in the past century have coincided with large changes in Western economies, making it difficult to disentangle them.

To overcome this obstacle, we study the long course of history, specifically examining periods with limited economic growth but substantial fluctuations in birth rates. Focusing on Amsterdam and Paris, for which we can reconstitute housing costs and demographic rates for almost 500 years, we test whether shocks to birth rates predict changes in rents or house prices decades later.

Existing research generally assumes that age determines the demand for housing space: young children contribute minimally, but demand for housing space surges when individuals reach early adulthood and leave their parental house. We confirm that this also holds historically, but that it generally has a limited impact on rental prices.

Demand for homeownership is different from demand for housing space.

The main novelty in our paper is that we also examine the timing of home purchases and whether this results in differential responses between house and rent prices. In most cultures and time periods, individuals typically begin buying a property when they are ready to settle down, often after marriage or its modern equivalents. Many people rent property before becoming homeowners, indicating that demand for housing space and demand for homeownership are different. As large cohorts start to settle, a large group of people enters the housing market to buy homes. However, typically only a few percent of properties are available for sale annually, implying a large cohort could significantly bid up house prices unless investors or existing homeowners supply more properties to the market.

Our findings indicate that this only happens to a limited extent, at least over time horizons of several years. We find that birth rates today strongly predict house price increases around 30 years later, but they have no impact on sales. Conversely, at older ages, current birth rates negatively predict house prices by the time this cohort begins to pass away or move to old-age homes. Overall, past birth rates can explain about one-fifth to one-sixth of the swings in house prices over the past centuries. This effect operates independently of other economic variables that might impact house prices.

Both today and in history, the house price effect dominates.

In essence, the impact of demographics on house prices does not arise from changes in the demand for housing space, but rather from the desire of birth cohorts to enter and exit homeownership at similar ages, while other actors in the housing market barely respond to this. While we can pinpoint these effects precisely with historical data, as birth rates and death rates were generally trendless before the late 19th century, we find very similar effects in modern OECD countries in the past decades: demographics have a large impact on house prices but not on rents.

The main takeaway from our findings is that shifts in population and its age structure are key determinants of house prices but have much less effect on rents. Countries where large cohorts will enter their home-buying years in the coming decade, such as India, may see rapidly appreciating prices, whereas those experiencing large generations retiring and passing away may see the opposite effect.