^{page 40 note (1)} See National Board for Prices and Incomes, Report no. 22, Rate of interest on building society mortgages, page 11.

^{page 40 note (2)} Of the 174,000 private new houses and flats constructed in 1970 in the United Kingdom 133,000 were purchased with building society mortgages, i.e. 76 per cent. For 1968 and 1969 the figure was 69 per cent. Sources: Financial Statistics, no. 108, table 67, and Monthly Digest of Statistics, no. 304, April 1971, table 109.

^{page 42 note (1)} Given these four major flows D, W, A and R and an approximation for IC, the remaining major element of table 3(ΔL) can be approximated by an accounting identity.

^{page 42 note (2)} See e.g. M. J. Hamburger, ‘Household demand for financial assets’, Econometrica, vol. 36, no. 1, January 1968; A. R. Latter and L. D. D. Price, ‘Substitutions between assets with fixed capital values’, mimeo., Bank of England Economic Intelligence Department.

^{page 44 note (1)} Most of these variables were suggested by the delegation of the Building Societies Association when cross-examined by the Committee on the Working of the Monetary System (Radcliffe Committee)—see Minutes of Evidence, Questions 7308, 7309, 7352.

^{page 44 note (2)} As equation (3) in the text above shows, the long-run coefficient is readily obtainable by dividing the estimated coefficient of the independent variable by the value of the adjustment coefficient (which is, as above, the complement of the coefficient attached to the lagged dependent term).

^{page 44 note (3)} Using the Durbin-Watson test in the usual fashion for each equation, and ignoring the problems of simultaneity and lags in the model, the null hypothesis (no serial correlation) is not rejected. However, it is established that the test is not strictly applicable to equations involving the presence of a lagged dependent variable, see M. Nerlove and K. F. Wallis, ‘Use of the Durbin-Watson statistic in inappropriate situations’, Econometrica, vol. 34, no. 1, January 1966.

^{page 45 note (1)} Stock market variables, the change in real disposable income, and the level of new mortgage advances (the latter suggested by the idea that house purchasers use their building society balances to finance down-payments) proved unsuccess ful. Inclusion of the lagged value of real withdrawals gave rise to variable results.

^{page 45 note (2)} The situation is different in the United States, where the institutional difficulties of changing mortgage rates are much less as Savings and Loan Associations alter the mortgage rate only for new lending, without recontracting the terms of all existing mortgages.

^{page 45 note (3)} To this extent our results confirm the view that ‘On the demand side, experience has indicated that the volume of mortgage applications is influenced much more by rising standards of living and capacity to pay than by the actual rate of interest charged’, Chairman's comment, Building Society Affairs, no. 54, March 1967.

^{page 46 note (1)} This was again suggested by representatives of the Building Societies Association in evidence to the Radcliffe Committee (Committee on the Working of the Monetary System, Minutes of Evidence, Question 7410).

^{page 46 note (2)} Here, the ‘slowness of adjustment’ is partly an artefact of statistical definition in that, whereas i is the recommended Building Societies Association share rate, m is the actual average mortgage rate.

^{page 46 note (3)} K* will then change from 0 to −1.33.

^{page 46 note (4)} This would, in turn, generate an increased net inflow of £43.8 million in the long run according to the receipts and withdrawals equations, (1) and (2) of table 4.

^{page 48 note (1)} In practice the values of the variables to be forecast are not significantly sensitive to reasonable error in L*, K*, which could in a real forecast situation alternatively be ‘guesstimated’ or approximated from forecasts of other items in the balance sheet.

^{page 50 note (1)} The estimation, by instrumental variables, used all 59 observations available.

^{page 50 note (2)} Inclusion of an expected equity price change term did not help matters.