Chapter 5 cited the UK mortgage market in the 1970s, and early 1980s, as a prime example of disequilibrium rationing. The existence of a mortgage cartel until 1983, and periods of negative real interest rates, led to mortgage queues and the use of non-price rationing mechanisms. These are the results we would expect if non-price terms of mortgage contracts were not being adjusted to clear the market (Drake & Holmes 1997). Thus the UK provides an ideal setting for the study of disequilibrium rationing, and financial deregulation and its effects.
Early UK research used time series data and generally incorporated dummy variables to represent regime changes, or variations in the severity of credit rationing (Nellis & Longbottom 1981; O'herlihy & Spencer 1972; Hendry & Anderson 1977). Some work used continuous proxy measures of the extent of rationing, such as the mean loan-to-value ratio (Ostas & Zahn 1975; Wilcox 1985), or the loan-to-income ratio (Nellis & Thom 1983). Studies using such continuous measures test for the presence of equilibrium rationing (see Holmes 1993). Rationing variables were typically statistically significant, and possessed the expected signs.1 However, Nellis & Thom (1983) and Holmes (1993) did not find any evidence of disequilibrium rationing. For example, coefficients on differences in the loan-to-value ratio between periods where not significantly different from zero.
One problem with most of the early studies of mortgage rationing was their inability to incorporate a major shift in regime, such as financial deregulation (Meen 1990). Meen (1989, 1990) follows some previous authors (Wilcox 1985; Hall & Urwin 1989) in using a direct measure of rationing where the demand function is implied by the structural equations of a supply and demand model. The extent of rationing was found by deducting estimated mortgage demand from observed supply. The empirical estimates indicated that mortgage rationing ceased to be a major problem for the UK after the middle of 1980, a result consistent with the findings of Wilcox (1985), and Hall & Urwin (1989). Excess demand and rationing was certainly apparent for earlier periods.
Meen found a large and statistically significant coefficient on the nominal mortgage interest rate in housing/mortgage demand equations. It is also possible to test for the appropriateness of a specification incorporating the user costs of owner occupation, and therefore the relevant interest rate specification (real or nominal) by introducing the nominal mortgage interest rate with some measure of expected house price inflation as separate variables. If the user cost is the appropriate specification then there should be no statistically significant difference between the coefficients on the interest rate and expected house price inflation, that is, they can reasonably be combined. A variety of tests of expectation formation mechanisms suggested that housing demand equations should not purely rely upon a real interest rate or user cost specification (Meen 1990). This issue concerned the tilting of mortgage payments discussed in Chapter 4. However, the tilt is not entirely separate from mortgage credit rationing. Restrictions on mortgage lending can exacerbate the tilt, also mortgage maximising households who are rationed will respond to nominal interest rate changes.
Muellbauer & Murphy (1997) estimated housing demand equations for the UK covering the period 1957 to 1994 and found that real interest rates were more relevant after financial deregulation. If households are at a corner solution with their demand for housing/mortgage debt then real interest rates will not be relevant. Muellbauer & Murphy also argued that expected income growth would reduce housing/mortgage demand under rationing as this was used to finance increases in non-housing consumption, a result borne out by the econometric estimation. The role of the nominal mortgage interest rate in econometric estimation of mortgage demand was also reinforced by the results of pooled time series/cross-section research (Leece 1995a, b, and 2000a). The tilting of mortgage payments, liquidity constraints and other capital market imperfections mean that nominal cash flows are an important consideration for some borrowers. This finding was seen to apply even in a financially deregulated environment, where mortgage contract design may still not be optimal (Leece 2000a).
The work of Ortalo-Magne & Rady (1998, 1999, 2002) demonstrated a correspondence between the outcomes of their consumption driven theory of housing demand and the aggregate behaviour of the UK housing market, pre- and post-financial deregulation. Ortalo-Magne &Rady (1999) estimate the relative contributions of income shocks or financial deregulation to the housing boom (1982 to 1989) and the bust (1990 to 1993). Their theoretical model, discussed in Chapter 2, indicated that rising income caused increases in house prices, with owner occupation among young households falling during the transition to equilibrium. However, this is not what occurred during the 1980s when owner occupancy rates among young households increased. Thus, the authors deduce that financial liberalisation allowed more young households to enter owner occupation.
Was this article helpful?