Hyman Minsky has become much more widely known since his death in 1996 because his ideas seemed to offer a framework for understanding the great financial crisis – the phrase ‘minsky moment’ has entered the language.
Homer Hoyt, who died in 1984, on the other hand is almost unknown outside the urban geography, urban planning, real estate economics field. However, where by contrast, and only in America (and in my house) he is somewhat of a revered figure in the small circle that still read his work, having made several major theoretical contributions.
Yet the ideas of each of them enormously enrich the ideas of the other – and offer an insight into what might be termed a Hoyt/Minsky process on how asset prices and credit, the physical economy and the monetary economy interact to drive financial stability/instability. Though Hoyt’s focus was on land/real estate, with adjustments these concepts are adaptable to a broader categoryof physical assets.
You can read lots about Minksy on the web, a good starting point are Randy Wray’s and Steve Keen’s recent primers, and I guess people would’t be coming to this page unless they knew a little of Minksy’s ideas.
John Monroe Illustrating the Hoyt real estate boom bust cycle
Hoyt’s personal story is like many of those (like Fisher) who were personally burnt by the Great Depression. During the 1920s Hoyt moved to Chicago, we has a realtor’s and had taught law and wanted to get in on the Chicago real estate boom of the 1920s. Following the crash of 29 and subsequent depression as well as being personally stung his mind turned to how real esate values are related to economic cycles. He became interested in land economics and became a doctoral student at the University of Chicago during its golden age. His doctoral thesis published in 1933 ‘ One Hundred Years of Land Values in Chicago‘ is of interest in this post.
This informed Hoyt’s later much more famous work on how cities grow, both physically and economically. I think these ideas, combined with modern non-neoclassical economic theory, have the potential for a ‘Cantillon‘ like understanding of the relationship between the physical economiy, geography and economic growth, only for the modern capitalist economy, rather than the pre-industrial capitalist one. Bold claim -and one to expand on in future – but to begin to understand this connection we have to go back to the genesis of Hoyt’s ideas.
A dissertation with the title of ‘ One Hundred Years of Land Values in Chicago’ would sound rather dull and institutionalist, but dull it is not and although institutionalist it is combined with theory and is a classic in how the two can and must inform each other.
The thesis had an immediate impact in puncturing the idea that property prices must always rise, as it outlined a series of five major booms and busts between 1830 and 1933. As Hoyt in his characteristic modesty put it.
The reader may wonder whether a consideration of the unique combination of events that produced Chicago will lead to the formulation of any principles of universal validity…In the history of a city there are elements similar to that found in the biography of a man. An extraordinary combination of hereditary factors, likely never again to be exactly repeated, placed in a historical situation that is also a unique complex of men and events, produces a type of human behavior which may not be duplicated. …A study of an entire city during the whole period of its growth, however, discloses a vision that might escape the glance of one whose horizon was limited by his precinct. The broad sweep of the events of a century reveals recurring cycles in the growth of Chicago in which general moods or similar historical situations are to a certain extent repeated. …It is possible that this long-run study of one city will lead to the discovery of factors that are characteristic of real estate activity in other cities, if allowances are made for inevitable differences in local histories and local environments (Authors Preface)
In chapter 5 – page 368 on – he drew together his 5 case studies and looked for common themes. And before you leap on a high horse and say – hey he’s talking about population and not about credit, he goes on to talk about the interrelation of population and credit – and as I go on to look at Hoyt’s ideas need modification outside of the context of a city’s physical expansion being driven mainly by in-migration.
The causes of land booms in American urban sites in the past century could usually be traced to factors which led speculators to expect an extraordinary increase in the population of the locality within a relatively short time, or the expectation of its development for commercial and industrial purposes. The anticipated population growth was itself due to deep-seated forces, operating over the entire area of Western industrial civilization….
necessary to the population growth of an urban center is the growth of its factories, transportation lines, banks, wholesale houses, and stores, which in turn depend upon the extent of its trading hinterland and its advantages for manufacturing plants, that land speculators seldom fail to stress these factors in describing the possibilities of further growth and the rise of land values of any city….
This increase in the rate of population growth was one of the factors that led to an increase in rents, building activity, and subdivision activity, each of which in turn was carried to speculative excess, and each of which interacted upon the other and upon land values to generate and maintain the boom psychology. (p368 -369)
He called this the ‘real estate cycle‘
In four of these cycles population increases led land prices – though of course the study preceded by half a century modern methods of analysing leads and lags. The one exception was the cycle ending 1879 where severe agricultural depression led to an influx from the countryside to Chicago, by contrast during the Great Depression the flow was to the West Coast.
Today we are more likely to consider wider issues of household growth rather than just population growth. The Chicago model works best for cities on virgin land or formed from agricultural inmigration (such as China). For more mature cities, as in America, Europe and Australasia today we also have to consider issues that lead to household formation in addition to population immigration, such as ageing (reducing turnover of properties), ups and downs in birth rates and deferred cohabitation, and divorce (increasing turnover and increasing household formation) . For example increased lifespan alone will require cities to grow to service the same population (in theory without any net inmigration) as properties will be occupied across their lifespan by fewer households. We also have to acknowledge that household formation is driven by economic growth as it requires sufficient income to form a home and sufficient taxes or expectations of profits to build them. Allowing for this consider Hoyt’s conclusion.
All five movements, Figure 90 shows, were characterized on their upswing by rapid increases in population, feverish building operations, and a hectic land boom in which land values increased from twofold to tenfold in a few years and their downswing by widespread declines in rents and foreclosures on a large scale which reduced land values 50 per cent or more from the peak levels and which brought building operations almost to a standstill. (p372)
He fleshed this out in the rest of this chapter into a theory of the real estate cycle – a summary of the chapter is set out below. I have highlighted the key sections where there is a supply shortage or oversupply and reaction in the monetary economy.
A Case History of Five Major Booms and Busts 1830-1933
1. Machine techniques, production methods improved
2. Population begins to spurt up
3. Shortage of housing, office & commercial space first felt
4. Rents begin to rise.
5. Selling prices of old buildings begin to advance
6. Vacant lot purchases begin to rise
7. Rate of new construction begins to rise sharply
8. Credit eases to stimulate volume of new building
9. Rapid growth of population projected far into the future
10. Prices of tracts near settled areas advance rapidly to peak.
11. Large tracts subdivided beyond needs of immediate development
12. Lavish public expenditures
13. Rate of population growth falls off
14. Vacancies reappear
15. Rise in rents slackens
16. Volume of building construction at peak.
17. Asking prices of land advance in face of fewer land sales
18. Financial institutions continue loans on peak values in face of lessened construction
19. Holders of 2nd mortgages begin to foreclose with faith in 1st mortgages
20. Stock market crash
21. Unemployment mounts to peak; wages down
22. Increased movement of population to small city or farm; doubling up in city
23. Vacancies mount to peak in houses, apartments, offices, stores; industrial rents down
24. Interest charges high in proportion to net rents
25. Taxes high in proportion to net rents
26. Second mortgage holders wiped out in flood of first mortgage foreclosures
27. Bank failures mount; loaded with real-estate “frozen assets.”
28. Volume of new building at bottom
29. Subdividing stopped; most vacant land not salable at any price
30. Construction costs at lowest pointsource: Homer Hoyt: One Hundred Years of Land Values in Chicago, Copyright University of Chicago Press, 1933
In a key passage Hoyt expanded on a key insight on Henry George nearly 40 years earlier
The supply of houses cannot be immediately increased. The new arrivals ..required housing facilities. They could not ship their old homes to their new abode as they could transport food and furniture. A mere shift in population from one place to another therefore increases the aggregate demand for new residential buildings, because the vacant space left behind cannot be transferred to the city to which the people are flocking. (p376)
In more modern terms we would talk about an increase in aggregate demand within a housing market area, caused by the interrelationship of economic growth in that market area, with effects on household formation, net turnover of properties in that market area and effects on inward and outward migration. However the fact that ‘the supply of houses cannot be immediately increased’ is the same. Land, and the property built on it, have a number of unique aspects of a factor of production, I will focus on just two for the preasent argument:
- Firstly it is in its virgin state, like water and air it is a free good. It is only when it becomes scarce, from the combination of property enclosure, population density and the advantages of resources/geographical location that that land enjoys that it has a scarcity, and hence a price at all.
- Secondly property development takes time – it cannot be immediately expanded – you have to buy it, get consent, build it and sell it to a market which will require time to get credit and where there is a chain sell their own properties.
And if there are any New Keynsians out there this is not an issue of price ‘stickiness’ it is a matter of the maths of dynamic disequilbrium- time alone is enough. By the time the property is bought and sold the market has moved on. It is a matter of the mismatch of the monetary circuit and the physical circuit of the economy.
When the monetary cicuit and physical curcuit are aligned it is like the San Andreas fault in calm times, pricing decisions based on cost plus pricing is easy, the friction between the two circuits prevents a sharp change in relative prices, when the pressure builds however, such as from a lag in housing supply, then pressure builds and creates the setting for market catasrophe.Key readings on this interaction from Benassy and Laidler [Nick Rowe btw is largely Laidler’s ideas in blog form].
On the rise in disproportionate rise in credit that the disproportionate increase on asset prices causes Hoyt says:
The building boom is stimulated and sustained by a liberal supply of capital made available by the expansion of credit institutions and attracted into new construction by the high profits made in such projects and by the high net yields of existing buildings. Financial institutions play an important role in this mania for building. …almost the whole amount required to erect the building and even to buy the lot could be secured in the heyday of these booms by loans on first and second mortgages (‘shoestring mortgages’ – as Hoyt calls them). (p 383 & 386))
At this phase of the real estate cycle, the rapid rate of increase…that has recently taken place is projected far into the future in the rosy calculations that are broadcast by real estate men. A city that will surpass in size any metropolis the world has ever known before is erected in these speculative dreams, and facts and figures are collected by business men of the community and by “distinguished scholars” to buttress these “castles in Spain” and to make them seem tangible to the lay mind (p388)
For Hoyt the market falls off because the financial speculation outpaces the physical market, population growth falls off as house prices outpace earnings, and, after the delay discussed abov,e a flood of new properties come on stream forcing prices down.
The holders of heavily mortgaged properties, who find themselves called upon to meet prepayments upon both first and second mortgages as well as the interest charges, begin to get into difficulties with the holders of the junior incumberances as soon as the peak income declines slightly or heavy prepayment falls…Up to this time there has been no major recession in general business activity. The end of the general period of prosperity in all lines is approaching, however, and on some red letter day…a crash on the Stock Market shatters the dream of never ending profits. ..The general slackening of industrial activity which continues after the stock-market crash, however, begins to wear down land values by a process of attrition. Within a year after the onset of the depression, increasing unemployment and reduced wages have sapped the public buying power….The reduction in the margin between income and operating expenses has now proceeded to the point where there is not even enough left to pay the interest charges on the first mortgages. The second mortgage holders, who were the first to foreclose, are themselves wiped out in the flood of foreclosures of first mortgages…All the factors now operate to depress real estate values as they operated to elevate them before….In this situation the financial institutions reverse their liberal lending policy of the earlier period. In most cases the banking power is badly crippled…
From this account it might be inferred that the real estate cycle automatically repeats itself. Such is by no means the case. According to the view presented, the cycle has been generated largely by a sudden and unexpected increase in population which was in turn due to a rush to take advantage of economic opportunities. If this theory is correct, then the recurrence of land booms in Chicago in the future will depend on the expansion of industrial opportunities which attract a sudden accession of population. (p398-401)
The sharp eyed may see this as a fleshing out of the thesis of set out by Simon Kuznets in 1930 in Kuznets S. Secular Movements in Production and Prices. Their Nature and their Bearing upon Cyclical Fluctuations. Boston: Houghton Mifflin, 1930. A kuznets cycle as we call it today – of 15-25 years. Kunzets is not mention in the references of Hoyt’s thesis – which doesn’t extend much beyond Chicago. Also Kuznets work was not known for many years, and it was not publicised in Schumpeter’s work on business cycles because of the considerable rivarly between Schumpter andKuznets. So to be fair we might call this an independent discovery and talk of a Kuznets/Hoyt cycle. (note Kuznets remarked dryly that the whole ofSchumpeter’s story of economic growth story boils down to the assumption that Schumpeter’s heroes (and heroines), entrepreneurs, are getting tired about every fifty years, whilst Schumpeter is a hero of mine as well, Kuznets was right that entrepreneurs are endogenous to cycles not exogenous).
It is here that the links to Minsky should become obvious, and his distinction between hedge, speculative and ponzi positions.
- In a Hedge position, expected positive cash flow from the investment is sufficient to make all debt payments as they are due, including both interest and principle on the loan.
- A Speculative position is one in which the expected cash flow from the investment is sufficient to make interest payments, but the principle has to be rolled-over. It is speculative as the positive cash flows must increase, or an asset must be sold on to cover the principle payment
- a Ponzi position is one in which even interest payments cannot be met, so the debtor must borrow more to pay interest (the outstanding loan balance on the creditors books grows by the interest due each period). Speculative positions will turn into Ponzi positions if positive cash flow falls, or if interest rates rise. It is a gamble on rising asset prices. Ponzi positions are fragile and risky as default is avoided as long as the creditor allows the loan balance sheet entry to keep growing, and the debtor will keep borrowing, at almost any interest rate, if their only concern is to stay above water. At some point beyond forbearance the creditor will cut losses by forcing foreclosure.
In Minksys famous thesis stability creates the climate for instability the mirage of stability, moderation and rising profits/asset prices, leads an in increase in speculative and ponzi positions – why innovate when there are lots of easy money and easy outlets for that money?
Specifically in terms of real estates and other assets that attract economic rent the valuation of that rent or imputed rent forms the basis of the valuation of the ‘fundamentals’ on a property loan, and whether the loan is ‘beyond fumandentals’.
Of course asset price driven booms and busts can be driven by any asset, such as in the dotcom and tulip booms, but in all such booms real estate plays a considerable part and in many provides a leading role. (see Fred Harrisson’s books for a detailed overview of the Hoyt/Kuznets cycle in history and in the 2008 crash, Mason Gaffney has also covered the continuing relevance of Hoyts ideas in some detail. )
Minskys ideas can be enriched by Hoyt/Kuznets as they help explain specifically what it is about physical asset prices and associated credit which drives the value/money mismatches that Minky talks about. Also Hoyt/Kuznets can be expanded by an appreciation of Minksy’s ideas on credit and disequilbrium and the broader plug-in he provides to both the disequilbrium economics of Keynes and Fishers ideas of debt-deflation. It also helps enlighten a debate between Randall Wray and holders of the monetary circuit theory (Sorry the next few paras on this are wonkish)
As summarised by Da Costa
As liquidity preference increases [that is the preference for holding assets as opposed to money], asset prices fall, causing a decline in physical asset output and, consequently, a decrease in the ex post spendings and income flows. In the opinion of [post keynsian critics of circuitism] authors, the divergence is basically because the circuit theory followers have their focus on the money flow created by credit, whereas [others] stress money as a balance that has to be held, because the future is uncertain.
The money balance view being essentially the same as Laidler’s concept of money as a ‘buffer stock’. The buffer stock theory is a good one however as asset prices rise we get the reverse process a decline in preference for holding idle money stocks and balances and an increased demand for credit to fund physical expansion. We need to make a distinction between the short run decisions made by entrepreneurs seeing opportunities in monetary/physical disequilibrium prices, and how medium term decisions which effect the price of money and sectoral balances by governments and central banks set the context for those entrepreneurial decisions – one is not reducible to the other nor even to the conventional and rather dated micro/macro distinction.
Of course both Minksy and Hoyt presented verbal models, systems models, and the big breakthrough of the last 15 years has been the development of mathematical stock-flow consistent (SFC) models. Will this extend to city modelling – overthrowing the general equilibrium Muth/MIlls/Alonso types models (ironically given the events of the last couple of weeks it was for work within this framework that Krugman got his noble prize – not macroeconomics – all at once ‘I can tell’ – and even more ironically these equilibrium models are assuma ptoltolemaic land surface of perfects circles and even expansion – one which Hoyt famously criticised), in a way that explains polycentric uneven urbanisation, Hoyt like corridor expansion, fringe belts, and the effects of the economics of agglomeration (increasing’/constant returns – a non neoclassical assumption – see my Why a City Scales Like an Elephant– which these old school models can’t).
This would be not trivial but fairly easy if we could use Python to plug in the outputs from SFC modelling programmes such as the emerging Minsky to the new wave of agent based models such as CitySim, Metronamica, Slueth and ArcGis ( with approaches such as LUCIS), but we cant because the SFC end lacks of a python interface or base as far as I can see, and lack of input/outputs of RDMS arrays stored in common XML formats for modelling cities like CityGML (or any arbitrary format using a web ont0logy language). With the right funding streams however this is well within the realms of possibility, and is an area where USGS, Erasmus, and the likes and IBM and Siemens have been very keen to fund urban modelling research proposals. This is back again to Hoyt’s follow up work on city growth economically and physically.
What we are talking here is a firm understand of physical assets undermining economic SFC modelling, and conversly this being used to model economic and geographic growth – the same parameter’s as Cantillon’s work .
Decisions, Decisions, Decisions 