What happens if newly-built houses cannot find buyers?
If only there was a simple formula for predicting market crashes, we'd all get rich. (Well, those who knew it would). But there isn't. Except that, just maybe, there might be when it comes to housing.
A tall claim - and one I don't intend betting money on. Even if I were legally entitled to give financial advice on the matter, I wouldn't. Still, it's an odd little piece of data and one which policymakers perhaps ought to keep in the back of their minds.
It shows a fairly consistent historical pattern that when the cost of mortgage repayment exceeds a third of average disposable income, house prices correct. That is to say, they fall.
As we know, they don't tend to fall softly. Another piece of data, from the OECD during the bubble, found that if house prices have risen more than 50pc during an upswing, they fall more than 30pc in the downturn.
Don't say nobody told you.
There is the rub. The future always looks obvious, looking back. But as the regulatory warnings say, the past is no guide to the future and, looking ahead, things are not at all obvious.
It is particularly hazy now because everything has changed so much since 2007. Even the recent past is another country, where they did things differently. Models based on what happened then, never mind economic policies, are likely to prove mistaken.
So one should not rely too much on the finding in the Budget report by consultants Indecon that a couple earning €84,000, living in the average Dublin house, would pay 30pc of their income on the mortgage - close to past danger levels. But nor should one ignore it completely.
Especially not the bit which says the figure would have been 17pc had they bought in 2012. On this measure, houses were cheap then, just as they were in 1997. Of course, being average, our couple could not have got a mortgage in 2012. Analysing the data is not much use if you don't have the dosh. It is, however, of some help in analysing the present.
The present is very peculiar, with no housing market worth the name. There are so few transactions, and credit is so constrained, that prices are not a convincing guide to conditions.
Sales are beginning to increase, but mainly in new houses. Purchases of those grew by 44pc from the same period of 2016 but from a low base and with two-thirds of the deals in the Dublin area.
Sales of existing homes were up a paltry 6pc, which means things were pretty much stalled outside greater Dublin. Among the many strange things about the times we live in is that, although there is this huge difference in the state of the market regionally, the spread in prices has never been narrower. That's probably down to lack of activity and affordability too.
It is difficult for people to reconcile all this with the other figure that, compared with 2007, Dublin prices are still 26pc lower and 30pc down elsewhere. In such dysfunctional conditions, where supply is nowhere near equilibrium with demand, it is tempting to avoid drawing any conclusions from the figures.
Yet we cannot just dismiss the dire possibility that only those households with earnings of more than €80,000 a year are in safe financial territory buying the average-priced Dublin home.
It has been argued that this may not be as dire as it might look, precisely because relatively few properties are involved. That may be the case where the health of the banks, or the wealth of most households, is concerned, but it is not the case with the current problem of insufficient construction.
The Indecon report, which was analysing the effects of the Help To Buy scheme (HTB), came to the conclusion that construction costs are the single biggest driver of rising prices. This was based on as extensive a survey of house builders as possible. More than half described "changes in cost of construction" as a significant, or very significant, factor behind price increases.
That was only one of a wide collection of causes cited by builders. The HTB scheme itself was described as significant or very significant by 45pc. Only 10pc thought it did not make some difference. The impact of revised loan-to-value mortgage rules and increased demand by first time buyers also figured prominently.
Indecon concludes that housing supply will be largely determined by the cost of construction compared to prevailing market prices. Other factors will be the availability of credit for contractors and builders' assessment of the sustainable level of effective demand.
The report does not go so far as to conclude that high-profile measures, such as the HTB and easier deposit rules, are pointless at best and harmful at worst - but there is good reason to think so.
The question left hanging in the air is what happens if present market prices do not prevail?
There are caveats around all these conclusions. The recovery is too recent to give reliable trends - including my own favourite repayment-to-income ratio - and survey responses must always be treated with caution. But we are left with the worrying thought that the cost of building an average dwelling is more than the average buyer can safely afford.
If so, any large increase in supply may not be met by customers who can afford it. Alternatively, prices may peak and fall in the next few years. Either way, the recovery in construction would stall.
The consequences of the housing shortage in an economy growing at 5pc, with the potential to perhaps continue growing at 3pc a year are beginning to be recognised, although they should have been recognised and prompted emergency responses long before this.
The economy's prospects have already been damaged but that will be only the start if the little progress which has been made in housing comes to an end.
Consultants should be called upon again, to find out if construction costs are really so out of line and if so, why?
The threat to prices comes from a possible rise in interest rates, amid growing alarm among central banks that their policies have not produced what they hoped - more spending and inflation - but rather the one thing they did not want, more debt and rising asset prices.
Led by the Federal Reserve, they are starting to move, but with all the nervousness of skaters venturing onto thin ice. Too big an increase too quickly and the global economy may crack in another crash. Take too long about it and the ice will just get thinner.
Their problem, and ours, is that it is already hard to see how a fall in asset prices can be avoided. Those prices are based on investors being happy with rental yields in low single figures, given that there is none at all on safe government bonds.
The best hope is that improving conditions in the global economy may prevent a fall turning into a crash. Ireland is vulnerable either way and time may well be running out.
The promised land of 2021 is just too far away. A lot needs to be done and it had better be done quickly.