Could a new type of mortgage reduce inequality?

There has been a long-term downward trend in the share and strength of labour in national income, which is depressing both demand and inflation. This has prompted ever more expansionary monetary policies. While understandable, indeed appropriate, within a short-term business cycle context, this has exacerbated longer-term trends, increasing inequality and financial distortions. Perhaps the most fundamental problem has been over-reliance on debt finance (leverage). We suggest measures, perhaps including government involvement, to raise the share of equity finance in housing markets; such reforms could be extended to other sectors of the economy. Nevertheless, such trends, though they have now lasted for some 35 years, are not immutable and might see a reverse in the course of the next 35 years. The recent Bank for International Settlements Annual Report (BIS 2014), claimed on page 7 that “Understanding the current global economic challenges requires a long-term perspective”, if we are to find a ‘new compass’ for setting macroeconomic policies. This column aims to provide such a perspective by focusing primarily upon one particular trend in developed countries, which is the trend decline in the adjusted wage share as a percentage of GDP in most developed countries since the end of the 1970s. This is shown for four main economies in Figure 1, and for a wider set of countries in Figure 2.

Figure 1. Adjusted wage share declining across developed markets

Source: European Commission AMECO database, Morgan Stanley Research

Figure 2. Developed-market adjusted wage shares (as % of GDP).

1970 1980 1990 2000 2010 2015F Germany 61.1 63.7 58.8 60.6 57.1 57.9 France 63 66.5 59.3 57.2 58.7 58.2 Italy 65.4 66.6 61.9 53.2 55.4 55.1 Spain 64.2 66.8 61.7 58.9 56.8 52.3 Canada 61 59.3 59.7 56.4 57.4 56.7 Australia 59.8 63.6 59.1 57.1 53.7 56.3 Denmark 60.2 62.3 59.3 56.4 59.5 57.7 Ireland 67.3 70 59.4 48.3 53.2 50.1 Greece 64.8 60.3 62.4 55.6 55 47.1 Norway 58.3 55.2 54 46.6 48.1 49.6 Netherlands 65.2 68.1 61.7 59.6 59.4 60.2 Source: European Commission AMECO database, Morgan Stanley Research

While the rate of decline varies from country to country, it nevertheless appears to be broadly common. This has reflected an initial fall and flattening in the rate of growth of real compensation per employee since the early 1980s, which has been continuing through the ups and downs of the economic cycle, and is again common to most developed countries; it cannot easily be attributed to short-term political or macroeconomic policies. This is shown for a variety of countries in Figure 3, and for a wider set of countries in Figure 4.

It is not our purpose here to try to explain why this has been happening; it is far too complex for us to tackle. Nevertheless, our preferred explanation is globalisation, and in particular the entry of the Asian, especially Chinese, labour force into the world’s trading economy. This has allowed businessmen to apply a credible threat of relocating the production of any good, and of most services, to anywhere else in the world. This has gone hand in hand with a decline in private sector unionisation almost everywhere, with causation going in both directions, and with a simultaneous decline in inequality between countries at a time when inequality within countries has been rising.

It is certainly possible that technical progress, in the shape for example of IT and robotics, has further weakened the share in output of labour, relative to capital and land (including natural resources, such as oil), but the measurement of this effect is fraught with difficulties. What we tend not to believe is that there is some immutable law whereby the return to capital must outstrip the overall growth rate of the country (r > g), as proposed by Piketty (2014); indeed, we intend to prepare a future paper that will argue that the above trends, observable over the last 35 years or so, are very likely to reverse over the next 35 years.

The macroeconomic consequences of a fall in the wage share

Be that as it may, what we do want to focus upon are the macroeconomic consequences of a trend decline in the wage share of GDP, and relatively low real wage growth, relative to returns to capital and land. Workers tend to be poorer, more liquidity-constrained and less likely to aim (or be able) to pass on wealth in inheritance to subsequent generations than the owners of capital and land. Hence they will, as a generality, have a higher (marginal) propensity to consume. So, the trend weakness in returns to labour will simultaneously tend to hold down consumption, output, and inflation (see, for example, the paper by Kumhof et al. 2013).

Figure 3. Real compensation growth weak across developed markets.

Source: European Commission AMECO database, Morgan Stanley Research

 Figure 4. Real compensation per employee (avg. annual % growth)

1970–1980 1980–1990 1990–2000 2000–2010 2010–2015 Germany 3.4% 0.5% 1.8% 0.6% 0.2% France 4.2% 0.8% 1.2% 0.9% 0.5% Italy 3.6% 1.2% 0.2% 0.7% -0.1% Spain 5.7% 1.3% 0.5% 0.6% 0.2% Canada 0.8% 1.1% 1.2% 0.6% 0.7% Australia 2.2% 0.4% 1.8% 0.1% 2.8% Denmark 2.2% 0.4% 1.8% 0.1% 2.8% Ireland 5.0% 2.3% 1.6% 2.6% -0.8% Greece 3.6% 0.0% 0.6% 0.9% -2.8% Norway 2.6% 1.9% 1.0% 0.7% 1.1% Netherlands 3.2% 0.5% 1.2% 1.3% 0.7% Source: European Commission AMECO database, Morgan Stanley Research

This weakness in consumption and output is likely to lead to fiscal expansion and rebalancing, whereby welfare and benefits rise, financed by higher taxes on the rich. Whether there is scope for further fiscal expansion, in view of heightened public sector debt ratios and prospective future claims on the public purse from an ageing population, and for a more aggressive rebalancing and fiscal redistribution, is yet another major topic that is too large and complex to be tackled here.

What we do instead note is that circumstances in which both real output and inflation are held down compared to target by the relative weakness of labour would seem, superficially, tailor-made to be rectified by expansionary monetary policy. Just as the share of labour has been declining since the 1980s, so has the level of real interest rates. This coincidence of a declining wage share and declining real interest rates is not, we believe, accidental.

From a short-term, business cycle viewpoint, a conjuncture of sluggish output growth and low inflation, surprising on the downside, should be met with and rectified by more expansionary monetary policy. But if one accepts the hypothesis that a (perhaps the) longer-term driver of such a conjuncture is the relative weakness of labour as a factor of production, then this short-term response is unhelpful, indeed somewhat counterproductive, in a longer-term context. Its main effect is to raise asset prices, and the relative value of land and capital, and thus benefit their owners, who are rich, rather than workers, who are poor. Hence, the trickle-down via the added consumption of the beneficiaries will be muted. More private sector capital expenditure would most likely benefit workers, e.g., by raising productivity and shifting the K/L ratio, but, alas, the empirical evidence shows capital expenditure to be notably interest-insensitive. Meanwhile, public sector investment is constrained by a variety of other factors, much to the dismay of Keynesians. What we are left with is the impact of monetary policy, whether conventional or unconventional, on housing and exchange rates. Movements in the latter are a zero-sum game globally; my depreciation is your appreciation.

Figure 5. Real interest rates negative across developed markets.

Source: Haver Analytics, Morgan Stanley Research

This leaves housing. The financing of residential and commercial property has become central to our banking systems in recent decades, whether directly via mortgages, or indirectly via loans to construction companies, loans collateralised on property, etc. And almost all such financing is done in debt form. The failure to reform housing financing modalities in the aftermath of the Global Crisis has been a sadly missed opportunity.

The decline in (long-term) real interest rates has been a major factor leading to a rise in housing and property prices, relative to labour incomes. It has, of course, benefitted those already on the housing ladder – the old and the rich. So, affordability has been on a trend decline, though less in the US than elsewhere.

Against this background, the bipartisan political incentive in the US to encourage more mortgage lending to the poorer, disadvantaged classes was entirely understandable. But it ended in the sub-prime crisis. Insofar as the main domestic transmission route of monetary policy to the real economy lies in the housing market, the authorities would appear to be caught in a dilemma. Either they encourage the young, the workers, and the poor to take on an unstable and excessive burden of debt, or they pump up all other asset prices further and further for less and less effect on the real economy, with a potentially growing risk of some future (disorderly?) reversal.

The current mantra is to constrain any incipient overheating and excess indebtedness in the housing market by macroprudential measures (e.g. the UK and Sweden), while pressing on with expansionary monetary policies to regain output and inflation targets. This raises several queries. First, will the macroprudential measures be pressed aggressively enough to work? But if they do, and the housing transmission channel is blocked, may not the resulting effects on other asset prices, including the exchange rate, have to be even more extreme (relative to labour incomes) and hence more distortionary and cause yet greater inequality?

There is, therefore, a longer-term, structural problem with monetary policy. As the BIS Annual Report (BIS 2014) noted on page 18:

“Policy does not lean against the booms but eases aggressively and persistently during busts. This induces a downward bias in interest rates and an upward bias in debt levels, which in turn makes it hard to raise rates without damaging the economy – a debt trap. Systemic financial crises do not become less frequent or intense, private and public debts continue to grow, the economy fails to climb onto a stronger sustainable path, and monetary and fiscal policies run out of ammunition. Over time, policies lose their effectiveness and may end up fostering the very conditions they seek to prevent. In this context, economists speak of ‘time inconsistency’: taken in isolation, policy steps may look compelling but, as a sequence, they lead policymakers astray.”

Whether this is true, yet, also for fiscal policy is a contentious issue, which we shall duck. Let us just state that policies of consciously allowing public sector debt ratios to rise further now are unlikely to prove acceptable in most countries. Having thus argued that expansionary monetary and fiscal policies have both largely ‘shot their bolt’, the BIS argues instead for “balance sheet repair and structural reforms”. Alas, we believe that its positive proposals are much less compelling than its criticisms of existing policies.

Whereas structural reforms and deleveraging are beneficial in the longer term, almost by definition, they are much easier to achieve during periods of fast growth, rather than the current sluggish expansion. Moreover, structural reform commonly involves removing the monopolistic rents of protected sectors, and thus initially is deflationary, prior to the subsequent expansion of output and productivity. The longer the stagnation, the more difficult it may be politically to introduce supply-side reforms, such as cutting subsidies. In the absence of specifics – and there are virtually none in the BIS Annual Report – a call for more structural reform and deleveraging is akin to an appeal to adeus ex machina.

On possible solutions

A better proposal is to be found in the book House of Debt by Mian and Sufi (2009), especially Chapter 12. The Global Crisis primarily impacted the poor, especially those subject to foreclosure, whereas the countervailing expansionary monetary policy mainly benefitted the rich, thereby worsening the longer-term trends already reported earlier in this column.

Their solution is for financial intermediaries to offer

“[A] shared-responsibility mortgage (SRM) [which] has two important differences: (1) the lender offers downside protection to the borrower, and (2) the borrower gives up 5 percent capital gain to the lender on the upside…” (p.192).

If housing price movements were independently and identically distributed, this would be a perfect solution. Unfortunately, they are not; they are strongly autocorrelated, with long periods of rising prices, sometimes culminating in a frenzy of sharp increases, interspersed with shorter periods of collapses in such prices. Figure 6. Housing (un-)affordability driven by post-2000 surge in house prices

Source: European Commission AMECO database, Haver Analytics, Morgan Stanley Research

The way that this would work is shown in the box below, which is mainly taken from Mian and Sufi (2014: 171–173).

The key difference between the SRM and typical mortgages is that the SRM provides downside protection to the owner if house prices fall. The owner’s mortgage payment schedule is linked to some form of local house price index: “For example, if her local house-price index is 100 when she buys the home, and falls by 30 percent by the end of her first year of ownership, [the owner’s] mortgage payment in her second year would decline 30 percent (…) Her amortization schedule would remain the same. As a result, even though she will make a lower payment, her mortgage balance goes down according to the original formula. This in effect means that [the owner] is given an automatic principal reduction when house prices in her area fall below her purchasing level. In our specific example, if [the owner’s] house-price index remains at 70 for the remaining twenty-nine years of her mortgage, she will have received a 30 percent forgiveness in principal by the end of her thirty years” (pp. 172–173). However, on average, house prices are expected to grow. At some point house prices will again likely exceed the original purchase price. Increasing gradually, once the house price index reaches that original level, the owner will be making full payments again.

During the upswing, and even perhaps especially in the final frenzy, home buyers may not be keen to share prospective capital gains with lenders, despite the self-insurance advantage. Some tentative attempts to introduce shared-equity mortgages in the UK apparently ran into consumer resistance during the long periods of rising house prices. Some complained that they had not understood the terms of the contract and had been mis-sold.

Perhaps more seriously, would a lender be prepared to offer downside price protection once the market had begun to crack? What rate would lenders have offered such protection in, say, Las Vegas in 2008? Given the pattern of autocorrelation on housing prices, we fear that too few mortgage buyers would seek such SRMs in upswings, and that no sellers would offer it, or not at feasible rates, during downswings.

Nevertheless we think that the idea of greater (self-)insurance via a larger equity element in the housing market is good. This was a key feature of the Help to Buy equity loan, which has now been extended to 2020. It should be possible to build on this. For example, the government could decree that all house purchases had to have at least, say, a 30% equity share, of which a minimum of 5% would have to belong to the purchasers, though both such numbers are, at this point, somewhat arbitrary and would need much more careful analysis in any practical application. Insofar as purchasers were unable to reach the 30% level themselves, they could either purchase an SRM in the market, or go to the government for equity funding. Of course, during downturns, the authorities would then be landed with almost all such housing finance, but:

1. This would be strongly countercyclical;
2. It would be politically (more) popular (than bank bailouts); and
3. Given all that we know about the housing market, it should be long-term profitable for taxpayers.

The media has, in general, been critical of the UK government’s Help to Buy equity loan scheme, but supportive of aggressively expansionary monetary policy. We reverse the argument – Help to Buy is part of the solution to a long-term structural problem which has been, in part, exacerbated by aggressively expansionary monetary policies.

However, the more fundamental problem is not those policies, but the financing context in which they operate, which gives advantages to debt over equity financing. The most egregious is the relative tax advantage of debt finance, and the ‘Holy Grail’ would be to equalise, or even shift, such advantages to the benefit of equity finance. Moreover, in a world with massive existing debt overhang, the transition to a much higher equity ratio can be very painful to existing shareholders, usually including top management, who are in a position to block any such move.

Among such changes to a more equity-financed world could be a change in the pattern of government finance. Thus, a move towards nominal income bonds should help to make government finance less procyclical, with less austerity during deflation and less temptation for Ministers of Finance to raid surpluses during booms. Of course, there are problems of revisions to – and even falsification of – data, but these could be handled.

Conclusions

There has been a long-term downward trend in the share and strength of labour in national income, depressing both demand and inflation. This has prompted ever more expansionary monetary policies. While understandable, indeed appropriate, within a short-term business cycle context, this has exacerbated longer-term trends, increasing inequality and financial distortions. Perhaps the most fundamental problem has been over-reliance on debt finance (leverage). We propose measures, involving government intervention, to raise the share of equity finance in housing markets; such reforms could be extended to other sectors of the economy.

Nevertheless, such trends, though they have now lasted for some 35 years, are not immutable. In a forthcoming note, we shall argue that they are likely to go into reverse in the course of the next 35 years, i.e., by 2050.

Disclaimer: Charles Goodhart is a Senior Economic Consultant to Morgan Stanley. Philipp Erfurth is a European Economist at Morgan Stanley. This article is not an offer to buy or sell any security/instruments or to participate in a trading strategy. For important current disclosures that pertain to Morgan Stanley, please refer to the disclosures regarding the issuer(s) that are the subject of this article on Morgan Stanley’s disclosure website. https://www.morganstanley.com/researchdisclosures.

References

Bank for International Settlements (2014), 84th Annual Report, 29 June.

Kumhof, M, R Ranciere, and P Winant (2013), “Inequality, Leverage and Crises: The Case of Endogenous Default”, IMF Working Paper 13/249.

Mian, K and A Sufi (2014), House of Debt, University of Chicago Press.

Piketty, T (2014), Capital in the Twenty-First Century, Harvard University Press.

Published in collaboration with voxEU

Authors: Charles Goodhart was the Norman Sosnow Professor of Banking and Finance at the London School of Economics until 2002; he is now an Emeritus Professor in the Financial Markets Group there. Philipp Erfurth is a European Economist at Morgan Stanley.

Image: A view of semi-detached homes in the Willesden neighbourhood of northwest London, August 15, 2014. REUTERS