LAND ECONOMICS

*Former Assistant Professor of City Planning, The Royal Institute of Technology, Stockholm, Sweden.

NEVER IN THE HISTORY of the United States was there a year like 1929. Economically it stands high above the preceding years, as well as those that followed, up to the Second World War. The gross national product (GNP) reached a peak six per cent higher than that of 1928 and nine per cent higher than that of 1930. Unemployment was reduced to 3.2 per cent.

Then, as at present, newspapers and magazines used to publish forecasts and prognoses for financial development but nobody would have dared to stand up in the spring of 1929 and tell the American people that within four years a quarter of the entire labor force of the nation would be out of work. Of course to predict is not to prevent, so it may have been just as well that the Great Depression of the thirties remained unpredicted. The Second World War was, indeed, foreseen and there were people who really knew how to avoid it and yet it came about just the same.

As we know, preparations for the attack against the free peoples of Western Europe and America were undertaken by governments pretending to be Friends of Peace. Thus, the Nazi and Japanese leaders managed to lull the free peoples into a sense of security. Contrary to this, the originators of the Great Depression did nothing to hide their actions and preparations. This must be so because the originators of the depression can hardly have been anyone else but the free peoples themselves.

Yet, it seems to be much easier to analyze the origin of the Second World War than to point out the never-hidden causes of the depression. Any reference to the Great Depression shows that it was exceptional but most comments on its origin assert that there was nothing exceptional about the preceding economic development.

The writer is not an economist and therefore he might be misinterpreting the facts but some economic data of the years immediately preceding 1929 seem to him to reveal very clearly an important cause of the subsequent depression.

For some time 1 have been working on programs for housing production for Swedish communities. With the excellent statistics now available in Sweden as well as in the United States on population, families and dwellings, it is easy to estimate the need for dwellings and rooms for a few years ahead. Any specialist in this field realizes that the rate of production is closely related to (1) the increase in the need for housing and (2) the demolition of dwellings that have served their time. A housing production substantially greater than the sum of the additional dwellings needed and the dwellings condemned cannot be economically feasible for a period of any length.

In the United States there was an extremely rapid growth of population from 1920 to 1926, especially with regard to the non-farm population. In 1920 there was a serious housing shortage due to the First World War. The non-farm population of some 75 million people increased by about 2 millions per year which gave rise to a demand for half a million new dwellings annually. Naturally, the building industry had to produce much more than that to eliminate the shortage represented by families sharing dwellings with other families and to provide for the demolition of the most dilapidated of the old houses.

The statistical data now available¹ show that in 1920 the construction of dwellings for the non-farm population amounted to only 250,000. Family statistics are not available for those years but we can assume that many more dwellings had to serve two families in 1921 than in 1920. In America it has been a creed, more or less, that every family should live in a separate dwelling. On the other hand, the demand for more than one dwelling for a family seems to have been very small, at least in the twenties when vacations were short or non-existent. This absence of elasticity of demand means that the need for more dwellings must have been extremely strong in 1921 and we should expect rents to go up and residential construction to increase. This is precisely what happened. The production of new dwellings rose in 1921 to 450,000, and thereafter to 720,000 to 870,000 to 890,000 and finally in 1925, to 940,000, as shown in the diagram (Chart 1). Evidently, 940,000 units were far too many for long-term. production. The demand for new homes plus a reasonable amount of demolition would have required 700,000 dwellings to be built each year. Sooner or later production had to relate itself to this figure. In fact, the shortage seems to have already vanished by about 1923 and it was later replaced by a surplus of unmarketable dwellings. Rents and prices began to drop from about 1925 onwards. The right moment to reduce housing production would have been in 1924 regarding the nation as a whole. (Of course, there must have been variations in both directions on the various local markets so it was really a net reduction of theproduction that was needed.) The 890,000 dwellings built in 1924 were nearly 200,000 too many as compared with a reasonable long-term program. And the 940,000 built in 1925 more than doubled the surplus.

Of course since the surplus was there, it became necessary to cut down production even more. But to reduce production from 940,000 units to something like 600,000 units in an important branch of industry such as home building is no easy matter. Above all. a reduction to less than two thirds of any branch of production did not fit in with the pattern of perpetual growth that dominated the general views of the day. Certainly, the demolition of houses could be carried out a few years in advance, so to speak. Houses do not wear out at a given moment and therefore the annual amount of demolition was, and always is, somewhat elastic. Therefore, it was possible to keep up production without visible signs of a surplus. But an increase in demolition during a certain period of time must necessarily mean fewer houses ready for demolition at a later date.

Thus, the problem of reducing the residential construction industry to its proper size was delayed. The reduced immigration quota in itself should have been a strong reason to reduce residential construction. Actually, housing production was cut down in 1926 but only to 850,000 units. This meant that the latent surplus was increased by another 150,000 dwellings. The production figure for 1927 was 810,000 which added a further 200,000 to the surplus. From then on, the annual increase in the nonfarm population dropped from two million to slightly less than 1.5 million, i.e., the real annual demand for new dwellings including replacement of demolished houses was now about 600,000 units only. On account of the latent surplus a much smaller production would certainly have sufficed for some years to come.

From the available statistics we can see clearly that in those years an alarming over-production of houses took place in the United States. But at the time nobody saw any danger at all-not even a Keynes or a Morgan. And even if there had been a prophet who had foreseen the future more clearly than the experts, it seems unlikely that building contractors and house builders would have paid any attention to his prophesies. Many economists had told the laymen that a new era had dawned, an era exempt from any serious slumps in its economic life.

Still, if people had perceived the danger of over-production-if it had been an established fact--what could possibly have been done? To reduce the annual residential construction from 810,000 units to say 500,000 was neither more easy nor less dangerous than it had been two years earlier to reduce it from 940,000 to 600,000. The opportunity for a sufficient gradual adjustment was lost in 1924, never to return. From 1925, or in any case from 1926, the causes of a crash were inherent in the general economic system and the coming events were more or less inevitable.

CHART I-THOUSANDS OF PERMANENT DWELLING

UNITS STARTED EACH YEAR 1920 THROUGH 1935 IN

NON-FARm AREAs IN THE UNITED STATES.

The still undetermined factor was the exact time of the catastrophe. This detail depended on the behavior of the customers and of the building contractors. In fact, they all behaved quite normally; they did what we would expect them to do: since rents and prices were decreasing and since the number of real estate foreclosures was increasing, it is very natural that the number of dwellings built was, as we have seen, slightly reduced each year. When, in spite of this constant reduction, it became still more difficult to market houses and when still more mortgages were foreclosed, the building contractors seem to have become more and more reluctant to start the construction of new houses.

Thus, in 1928 the number of dwellings built was 750,000, in 1929 it fell to 510,000 and in 1930 it fell furtherto 330,000. That is to say, the first really big slump occurred in 1929-after four or five years of steady over-production in spite of the gradually reduced production figures. But even in 1929 the slight downward trend had spread to the brick manufacturers-who must have found it increasingly difficult to market a product dependent so much on the residential production. The reduction in the number of dwellings built in 1929 meant a curtailment in one year of one third of the annual production of this branch. This also meant the dismissal of somewhere between 1.5 and 2 per cent of the entire non-farm labor force, a percentage corresponding to the normal annual increase in this labor force. It meant a reduced call for bricks, carpenters' tools, wood, trucks, Portland cement, iron pipes, baths, plumbing, fixtures, etc. It also reduced the purchasing power of those unemployed and perhaps of their friends and relatives who probably began to save their money in fear of unemployment. It meant that tens of thousands of building contractors were out of business and had to sell stock to avoid bankruptcy. It meant that all those involved in business must have realized that the boom could not last much longer. And yet the 510,000 units produced in 1929 did not mean that the surplus of dwellings was reduced in the least. Actually, the following year the annual increase in the non-farm population was only about 700,000 (corresponding to some 200,000 families). The situation on the labor market in late 1929 caused many people, who would otherwise have moved to the cities, to stay on the farms. Later on, the net immigration was replaced by a net emigration for the same reason, thus further reducing the growth of population in America. Although we know very little about the demolition of houses in those years, we may conclude that the amount of demolition must have been extremely small for two reasons: (1) most of the deteriorated residential houses had already been demolished and (2) the prospect of unemployment and economic depression probably made house owners consider that even an old house was better than none. just as demolition can be carried out in advance, it can also be postponed a few years if the situation so demands.

Thus, the real demand for new dwellings in 1930 can hardly have been much more than 200,000 units. Yet, as mentioned above, the number of new dwellings actually built in that year was 330,000. Apparently, the builders could not visualize a change as thorough and rapid as that which actually took place. Still, the reduction from 510,000 to 330,000 new dwellings must have caused a further increase in the number of unemployed in the building industry as well as in several allied industries. The psychological factors did the rest. Why buy a new car when times are bad? Why scrap an old car when the motor still works? A new suit is dispensable, at least for a year or two, especially when the wardrobe closet is full. Thus, even more people had to be dismissed and still fewer moved to the cities. In 1931 the increase in the non-farm population was about 400,000, corresponding to slightly more than 100,000 households. Residential production still lagged behind and 250,000 new dwellings were built. In 1932 there was actually a decrease in the non-farm population but 130,000 dwellings were built for this population. In many respects this was the worst year. The chain reaction of unemployment, decreasing demand and still more unemployment continued until this low-water mark was reached.

Actually, several chain reactions operated in parallel throughout these years: e.g., (1) unemployment in the residential construction branch, reduced influx of people to the cities, reduced demand for dwellings, still more unemployment for construction workers, etc.; (2) constructional unemployment, reduced demand for building material, unemployment in the building material branch; (3) constructional unemployment, reduced demand for durable consumer's goods, unemployment in the various durable goods production industries.

However, a chain reaction does not start from nothing. At the beginning of each chain there was always the surplus of dwellings. As we have seen, this surplus increased from 1924 to 1932. As long as there was a surplus of dwellings, there was of course little else to do but to cut down the production in proportion to the reduced requirements, thereby starting new downward chains. Thus, the existence of a surplus of dwellings seems to have been a predominant detrimental factor and perhaps the only one that would seriously endanger the entire economic life of the nation in the late twenties. At least some researchers in this field have come near to this conclusion. Thomas Wilson writes, ---The collapse occurred only because the development of under-consumption was accompanied by a declining demand for houses and a serious exhaustion of investment opportunities."²

Alvin H. Hansen says, ---Thedrastic decline" (in housing capital expenditures) "from 1928 to 1929 of $1.2 billions undoubtedly exerted a heavy downward pressure upon the whole economy and contributed largely to the general collapse which started in 1929."³ And he specifies that, "probably at no time in our history had we reached as complete a temporary saturation in building construction, including apartment houses, residences, office buildings and other commercial structures, as was the case in the late twenties."⁴

Miles L. Colean and Robinson Newcomb direct their interest a step further back in the process when they write about the changing demand for houses and buildings:

"War is thus a great unstabilizer. The basis of the war effect is the creation of a huge accumulation of demand, followed by a sudden release in which all the accumulated demand seeks to be effective at once. The resulting boom is certain to be followed by a market glut as demand is either satisfied or choked off by a rise in p~ices and then by a period of readjustment in which surplus supply is absorbed, prices decline, and producers reorganize themselves to a less frenetic level of activity."⁵

It is well known, e.g., from Haberler's extensive review of depression theories,⁶ that many authors have discussed various other mechanisms which may cause business slumps or contribute to such. And of course many factors operating at the same time determined the actual course of events and the severity of the Great Depression. What I would like to stress is, however, the leading part played by the existence of a surplus of dwellings. This is something that may occur again even in a planned economy. As a matter of fact, an over-production of dwellings and other buildings is about to occur now in Sweden as a very result of authoritative planning, as the present writer has recently shown in another article.⁷ An over-production in the construction branch would probably be enough to cause quite a noticeable slump even when the rest of the economic development is entirely sound. It should not be impossible to compute rather accurately how much the 1928 to 1929 decline in the construction industry contributed to the general collapseas Dr. Wassily Leontief has pointed out in private communication to the writer. Such a computation could be carried out with the help of the %nverted inputoutput matrix" presented by Evans and Hoffenberg,⁸ after the quantities having been valued in proper prices, of course.

The Great Depression, to the extent that it depended on the surplus of dwellings, could not possibly have been forestalled because of the lack of statistical information of the demographic development. In order to prevent the depression,measures should have been taken as early as in 1924. Today, we should have every opportunity, in America as well as in Sweden, for example, to prevent a business slump which might originate from a surplus of dwellings. All that is needed is that we pay due attention to the statistical information on population and residential construction and that we take a lesson from the events which occurred in the United States in the twenties and that we act in time.

1 United States Bureau of the Census, Historical Statistics of the United States, Colonial Times to 1957 (Washington D.C.: Government Printing Office, 1960). All statistical data in this article are taken from this document.

2 Thomas Wilson, Fluctuations in Income and Employment, 3rd edition, (London, England: Sir Isaak Pitman & Sons, Ltd., 1949), p. 156.

3 Alvin H. Hansen, Fiscal Policy and Business Cycles (New York, New York: W. W. Norton & Company, Inc., 1941), p. 57.

4 Alvin H. Hansen, Business Cycles and National Income (New York, New York: W. W. Norton & Company, Inc., 1951), pp. 46-47.

5 Miles L. Colean and Robinson Newcomb, Stabilizing Construction: The Record and Potential (New York, New York; Toronto, Canada. London, England: McGraw-Hill Book Company, Inc., 1952), p. 131.

6 Gottfried Haberler, Prosperity and Depression, 3rd. edition, (Geneva, Switzerland: League of Nations, 1941).

7 Carl O. Nordling, "Byggprognoser i överkant," Byggnadsindustrin, Vol. 36, No. 15, September 1966.

8 Duane W. Evans and Marvin Hoffenberg, "The Interindustry Relations Study for 1947," The Review of Economics and Statistics, May 1952, pp. 97-142.