RDP 2001-08: City Sizes, Housing Costs, and Wealth 4. Deregulation, Disinflation and Housing Wealth Dynamics

The previous sections set out what we believe to be a plausible partial explanation of the current distribution of relative housing prices across countries. If Australia had always had relatively high housing wealth, our story would end there. However, in the early 1980s, Australia's ratio of dwelling prices to income was not obviously different from those in other countries, and Australia's households had much lower debt than households in many other countries. As mentioned earlier, we attribute this to constraints imposed on dwelling prices by high inflation and financial regulation. However, the extent to which housing was tax-advantaged was also important, because that tended to dampen the effects of high inflation and regulation. In countries such as Australia, where government intervention in the housing market did not offset the effects of inflation very much, actual dwelling wealth was further from its desired ratio to income prior to deregulation. Therefore, after deregulation and disinflation, these countries had further to travel to reach their desired level of dwelling wealth. The market adjustment in Australia was particularly prolonged, as dwelling prices in the larger cities had to converge to a higher long-run equilibrium, as determined by Australia's urban structure.

When the financial sector is regulated, credit is rationed and households cannot borrow as much as they would like at the current interest rate.^[16] Once these constraints are removed, household indebtedness usually rises, as shown in Figures 5 and 6 for eight countries; the period where deregulation took place, marked by a grey band, is frequently followed by a rapid increase in indebtedness.^[17] By contrast, Canada has always had a fairly deregulated home loan market (Edey and Hviding 1995; Freedman 1998), and never experienced the rapid run-up in debt seen in Australia, New Zealand, Sweden or the UK, which reflected pent-up demand following the removal of restrictions.

Some of the increase in indebtedness may have been due to existing home owners withdrawing housing equity, for example by refinancing and increasing their mortgages or by taking out a home equity loan. Households could also withdraw equity indirectly, for example if a household that inherits a house that was owned outright sells it to a household that took out a mortgage to make the purchase, and either consumes the proceeds or channels them into non-housing assets. This would result in higher consumption but not necessarily higher dwelling prices. Still, the removal of credit constraints on mortgage borrowers enables some households who previously did not own a home to purchase one, and some existing home owners to upgrade to a better home. This should be expected to result in increased effective demand for owner-occupied housing and therefore upward pressure on dwelling prices.

The rapid build-up in household indebtedness, however, appears to be more closely associated with a reduction in inflation to low levels. Figures 5 and 6 show that the most rapid increases in debt have occurred in Australia, New Zealand and the UK, where disinflation has followed financial deregulation. This suggests that disinflation dominates deregulation as a precondition for rising household debt. It is also worth noting that in Australia, the full benefits of financial deregulation did not accrue to households until the 1990s, with financial intermediaries mainly directing their lending toward businesses in the 1980s (Stevens 1997).

Even with a deregulated financial sector, high inflation will still constrain household debt because some imperfections and information asymmetries remain. Mortgage contracts are usually based on regular repayments fixed in nominal terms. Given this, financial institutions will only lend to households as much as they can reasonably service on their current incomes; in Australia, most financial intermediaries impose a maximum loan size corresponding to a repayment-to-income ratio of around 30 per cent. It is the nominal interest rate, not the real rate, that determines this repayment ratio. Therefore the higher is inflation, and thus the higher are nominal rates, the more households are affected by this market imperfection and excluded from the home loan market (Lessard and Modigliani 1975; Stevens 1997). Disinflation will therefore unambiguously increase demand for owner-occupied housing, although this effect may have been dampened in countries such as the United States, where mortgage interest is deductible. Existing home owners will be able to upgrade, current renters will be more likely to be able to move into home ownership, and new households may form as an endogenous response to the reduced costs of mortgage finance.^[18] This increased the effective demand for housing, resulting in higher housing prices and a relatively static home-ownership rate.

Deregulation and disinflation certainly affected household debt, but it did not necessarily follow that dwelling wealth-income ratios increased in all countries. Growth in dwelling wealth after deregulation and disinflation largely reflects the difference between its actual level in the regulated period and its putative desired level. The relationship between the actual and desired level was determined by the extent to which distortions in the housing market offset the effects of high inflation and financial regulation.We have already seen that these policies were not uniform across countries (see Table 3). In the presence of credit constraints, mortgage interest deductibility and capital gains exemptions on the family home in the UK and US made dwellings very attractive relative to other investments. This may have worked to offset the effects of regulation and thus closer align actual dwelling wealth with its desired level. Furthermore, mortgage interest deductibility in the UK and US would have ameliorated the burden that high inflation placed on mortgage borrowers; Britten-Jones and McKibbin (1989) found that changes in mortgage interest deductibility have a much larger effect on the housing market than changes in income taxes. These factors combined to mean that the actual level of dwelling wealth under regulation was possibly closer to its desired level than was the case in Australia.^[19] That dwelling wealth-income ratios in the US and UK are currently around their 1980 level is consistent with this point (see Table 2).

Fewer housing market distortions in Australia and New Zealand made the relative constraint placed on dwelling wealth by high inflation and financial regulation more binding. These constraints seem to have disproportionately affected cities that would otherwise have had high housing costs. The suppression of this urban structure effect widened the gap between actual dwelling wealth and its desired level. It took financial deregulation and disinflation to release this effect, and since then dwelling prices have responded accordingly, with the largest increases seen in Sydney and more recently Melbourne. It is therefore possible that national average dwelling wealth has been able to rise to its long-run level, now that housing prices in these cities are no longer constrained by these regulations. This may explain why Australia's dwelling wealth-income ratio has increased relative to other countries, from around the international average to well above it.

On the other hand, if the combined effects of high inflation and financial regulation had kept housing prices artificially low, they may also have had an effect on the current composition of the dwellings stock. These constraints on purchase prices would have affected building costs very little, so most of the effect would have manifested in land prices. Although home buyers may have been constrained from paying as much as they would in a deregulated environment, the artificially low land prices might have allowed inframarginal home buyers to purchase a higher-quality home than they would if prices had been higher. In particular, they might have been able to purchase a home that used more land. This might go some way to explaining the greater prevalence of detached houses in the Australian housing stock which, as alluded to in Section 2.2, might indicate a more land-intensive component to past demand for new housing than occurred in other countries.

Because the stock of housing greatly exceeds the flow of new building, it takes a long time for the characteristics of the stock to adjust to structural change such as a new post-deregulation equilibrium price. Therefore although the equilibrium price of housing may have risen following these changes, the aggregate value of the housing stock might be above its long-run level because the composition of the housing stock is yet to adjust fully. Now that land prices are no longer held down by inflation and financial regulation, people will tend to choose less land-intensive housing than in the past, supporting the trend to medium-density housing, especially in the larger cities.

Footnotes

Duca and Rosenthal (1994) found that borrowing constraints lowered the US owner-occupation rate by around 8 percentage points, disproportionately affecting young households. [16]

Appendix B contains the sources we used to construct the dates of the period of financial deregulation in each country. [17]

Poterba (1984) suggests that the interaction between inflation and mortgage interest deductibility could have explained the strong growth in US house prices during the 1970s. However, demographic factors such as the entry of the baby-boom generation into its house-buying years have also been cited as a potential cause (Mankiw and Weil 1989). [18]

In the UK, some of this incentive would have been undermined by the existence of a large public-housing sector. Pent-up demand for dwellings was unleashed in the late 1980s when some public housing was privatised (Henley 1999). [19]