Shared equity schemes: evaluation reports

3. Housing market context

This chapter provides an overview of the housing market trends that influenced introduction of the three shared equity schemes. Rates of house price inflation and the affordability of owner occupation are shaped by a mix of cyclical developments and long-term structural changes. This chapter begins by summarising long-run house price movements before focusing in on housing market developments since 2001. It looks at housing market activity in the period to 2008, which saw the introduction of the Scottish Government's Low-cost Initiative for First-Time Buyers (LIFT) programme, which included the launch of OMSE and NSSE. It then details the subsequent adverse housing market conditions that led to the introduction of HtB before looking at market conditions since 2014-2015. This chapter also explores sales activity for the three schemes and how these have been positioned within the housing market, which is particularly important for understanding the role of HtB.

Trends in house prices over recent decades

After adjusting for inflation, real house prices in Scotland have moved in a gradual upward direction for much of the last 30 years (see figure 3 over the page). The main exceptions have been in the boom years from 2003 to 2008 when house prices in Scotland and the rest of the UK increased rapidly and the years from 2009 to 2014 when prices declined as the impacts of the global financial crisis (GFC) took full effect.

Source: Office of National Statistics (ONS) (2019) House Price Index accessed December 2019, adjusted by study team using GDP deflator, which is a measure of general inflation in the domestic economy. The ONS average price is a geometric mean and is closer to the median than the mean.
Note: Price data for some UK regions is only available from April 1992 or later.

Over the long run, real house price inflation in Scotland has been less volatile and house prices have remained closer to the underlying trend path than in several other UK regions, and in particular London. Between 1990 and 2019 real house prices in Scotland more than doubled whereas house prices in London more than trebled. Part of the reason for this is in the 5 years to 2018 real house prices in London increased at faster rate than at any point in the last three decades. This has contributed to a widening of the gap between Scottish and UK house prices. In October 2019 the average (geometric mean) mix adjusted house price for Scotland stood at £153,692, which was 34% below the UK average of £232,944. In October 1995 the comparable gap was 23%.

The gradual increase in real house prices, punctuated by a period of rapid increase followed by a period of downward drift, can also be seen in nominal house prices, including trends at different points of the house price distribution. Nominal house prices for properties at the lower quartile price point are often assumed to approximate entry level house prices. Figure 4 shows that in the five years to 2008, lower quartile prices increased by an average of 14% per annum. This was in line with median prices and only slightly lower than prices at the upper end of the market (15%).

Source: Scottish Government (2019) Open Access House Prices; Based on residential property transactions recorded by Registers of Scotland: statistics.gov.scot/data/house-sales-prices

Price data for different types of property is only available from 2003-04 and is not fully robust, mainly due to the proportions of dwellings that cannot be assigned to a specific property type. In 2018-19 10% of dwellings were unassigned, although the comparable proportions prior to 2016-17 are less than 3%, reflecting time lags in data availability, especially in respect on new build properties. However, it appears price movements vary little by dwelling type. Figure 5 shows that nominal house prices for each of the various dwelling types grew by upwards of 50% between 2003/4 and 2007-08 but thereafter grew by just 10-15% in the decade to 2018/19.

Source: Registers of Scotland (2019) Property Market Report 2018-19

Over the last three decades average prices for new build properties have exceeded those for resale properties. The Office of National Statistics (ONS) House Price Index reports on nominal house prices for new and resale properties. This data is only available from 2004 but it confirms that the gap between new build and resale prices persisted throughout the period to 2018 (see figure 6). At the peak of the boom in 2007 and the first half of 2008, the average (geometric mean) new build price was 24% higher the average resale house price. Since summer 2013 nominal house price inflation for new build properties has outpaced that for resale properties. However, it is not possible to ascertain whether this has been due to differences in the mix of products sold or to an increase in any price premium attached to new homes.

Source: ONS (2019) House Price Index accessed December 2019.

Housing market developments: 2001-2008

As already noted, house prices were rising sharply before the global financial crisis erupted. This development was driven by a combination of demand fundamentals, consumer confidence and supply constraints, although relative importance of the factors that underpin house price growth and volatility remain contested (Mulheirn, 2019; Meen, 2019; Bramley, 2018).

On the demand side, sustained economic growth in the decade to 2008 increasingly fed through into growth in employment, earnings, household disposable incomes, inward migration and household numbers, especially from 2003 to 2008. At the same time, low interest rates, mortgage market innovation and the increased willingness of lenders to grant large loan-to-income value ratios on new mortgages boosted the borrowing capacity of households. Taxation policies and poor returns on stocks and other assets also saw small scale investors turn to the housing market, often with the expectation the prices would continue to rise (Scottish Government, 2007 & 2010; Maclennan and O'Sullivan, 2008; Stephens, 2011).

Sources: NRS (2019) Household estimates: Scottish Government (2019) Housing completion statistics and Scottish Government (2019) Stock by tenure estimates.
Note: Net stock figures are derived from the annual change in dwelling stock estimates at March that are based on the NRS dwelling counts which are known to be subject to lags and other errors in data.

On the supply side, overall housing production in Scotland increased. House completions in all tenures increased by 11% from an annual average 22,239 in the 5 years to 2003 to an annual average of 25,064 in the 5 years to 2008. However, as the Scottish Government observed (2007), there was little evidence that this was an adequate supply response to rising house prices, which we estimate increased by 87% in real terms in the 5 years to January 2008 (see figure 3). There were also signs that net growth in the housing stock was potentially lagging behind net household growth in the 5 years to 2008 (see figure 7), although data limitations make it hard to reach firm conclusions. The constraints on new supply have been variously linked to land constraints, rising land prices, complexities in the planning system, labour shortages in the construction sector and the investment behaviour of private developers.^[6]

As the first decade of the 21st Century progressed, growing numbers of households, mainly younger households under the age of 35 years, became increasingly priced out of the house purchase market. Figure 8 shows that even before the financial crisis began, homeownership rates amongst younger households had fallen from 53% in 2001 to 46% in 2007. Over the same period the proportions of younger households renting privately increased from 17% to 27%, indicating that the growth of private renting was partly stimulated by demand from those priced out of the house purchase market.

Source: Scottish Government (2018) Scottish Household Survey 2018: accompanying tables.

UK Finance statistics also confirm that the surge in housing market activity before the financial recession took effect was progressively driven by demand from existing homeowners and small-scale investors as opposed to first time buyers, such that:

• In 2007 some 35,000 new loans were issued to first time buyers in Scotland, 15,000 fewer than in 2001.
• The proportions of new mortgages to support house purchase issued to first time buyers (as opposed to home movers) in Scotland dropped from 50% in 2001 to 32% in 2007.

Figure 9 reports on affordability ratios based on median house prices to gross median wage for people in full time work. It shows that affordability pressures rose sharply at the height of the housing boom, peaking in 2007 when the average prices were up to 5.9 times higher than the average salary. It also shows that affordability ratios based on lower quartile house prices to lower quartile earnings produced a very similar picture.

Sources: Registers of Scotland (RoS) median and lower quartile prices and ASHE median and lower quartile earnings.

The affordability ratios reported in figure 9 overstate the extent of affordability pressures. This is partly because first-time buyers are often dual earner households and partly because they make no allowance for interest rates and how this affects the ability of households to make mortgage payments. Figure 10, therefore compares the mortgage burden of first-time buyers for Scotland and the UK. This term refers to the share of household income the average first-time buyer has to spend on mortgage repayments. It shows that mortgage outgoings for first-time buyers in Scotland peaked at over 22% in 2007 but remained some 2% below the UK average.

Source: UK Finance (2019) Regulated Mortgage Survey Table RL1: First-time buyers, new mortgages and affordability, UK countries and regions '4-10-2019.

UK Finance figures used to calculate the mortgage burden also suggest that at the height of the boom between 2005 and 2007:

• Average gross incomes for first-time buyers in Scotland increased from £28,000 in May 2005 to £34,200 in June 2007. By way of comparison, ONS (2010) report that the gross household income of households in the UK at middle of the income distribution was £29,100 in 2008-09.
• The average mortgage deposit fluctuated within a couple of percentage points of 20% of the purchase price, just as they had in the years since 2001, but in cash terms the average deposit increased from £13,200 to over £19,000.
• In 2007 the average first-time buyer would have to save the equivalent of over six months of their gross income for a deposit of £19,000.

Despite the increase in the cash value of the deposits made by first time buyers, UK Finance figures indicate there was no clear upward trend in the age of first-time buyers. Instead, the average age of first-time buyers generally fluctuated around the age of 32 years. One probable reason for this was that rising numbers of first-time buyers drew on parental assistance instead of saving for a longer number of years to build up a deposit. Clarke (2011) reports that the proportions of first-time buyers in Scotland buying unassisted declined from 69% in 2005 to 60% in 2007 and 52% in 2008.

Overall, in the period up to early 2008, there was a growing gap between house prices and the capacity of households to pay. As increasing numbers of prospective first-time buyers were priced out of the market, the Scottish Government concluded that a policy response was required. OMSE and NSSE were therefore introduced to assist households on low to moderate income to enter homeownership. Essentially, both schemes were intended to enable households to overcome the financial barriers that prevented them from buying by reducing the initial costs of purchase.

Housing market developments to 2014

As the global financial crisis deepened, severe restrictions were placed on credit and the UK economy entered recession. The housing market also slumped as finance for both house purchase and new housing development became very difficult (Whitehead and Williams 2011). The evaporation of consumer confidence in the housing market and the sharp fall in real house prices (see figure 3) had an immediate and dramatic impact of the housing market. Between 2007 and 2009:

• Annual levels of house sale transactions fell from 154,403 to 69,032 (Registers of Scotland (RoS) statistics).
• Mortgage approvals more than halved from 99,270 to 45,980 and approvals to first-time buyers dropped from 34,980 to 17,580 as more people turned to private renting or opted to live with their parents for longer (figure 11).
• Private sector housing starts more than halved from 20,626 to 9,309 and housing completions followed suite. In 2012 there were just 9,998, down from 21,685 in 2007 (see figure 12).

Source: UK Finance, first time buyer and mover new mortgages and affordability statistics.

Source: Scottish Government new build statistics.

After continuing to decline over 2010-12, real house prices began to outpace inflation in 2013 and 2014, but by December 2014 real house prices remained 17% lower than in December 2007. Likewise, whilst there was an upturn in private sector completions in 2013-14, levels remained below those seen prior to 2008.

The combination of a fall in real house prices and low interest rates improved affordability as measured by price-to-earnings ratios and the cost of servicing the average first-time buyer mortgage. However, housing market activity remained weak and there was only modest recovery in the numbers of loans made to both first time buyers and existing owners moving home. One potential reason for this was that households were uncertain about the prospects for the housing market and were unwilling to risk buying. Another was the significant challenge prospective first-time buyers faced in raising a deposit. As figure 13 illustrates, loan-to-value (LTV) ratios dropped as lenders sought to limit their exposure to possible losses. As a consequence, first-time buyers typically required a deposit of 25%-30% during 2009-2014. The net result was that there was some increase in the numbers of prospective first-time buyers with sufficient income to service a mortgage but without the means to raise a deposit. This in turn contributed to the continued decline in the rates of younger households in the owner-occupied sector in 2009-2014 (see figure 8).

Source: UK Finance - first time buyer mortgage and affordability statistics.

The fragility of the housing market lead both the UK and Scottish Governments to introduce measures to support the housing market. In September 2008 the stamp duty threshold was temporarily increased from £125,000 to £175,000 and was then extended in 2010 to £250,000 for first-time buyers for a two-year period. Low interest rates were sustained, in part to ease mortgage payments. There was also an injection of funds to stimulate housing development. The Scottish Government, for example, worked with the National Housing Trust to develop mid-market rented homes.

Most significantly from the perspective of this study, HtB was launched in 2013. As a policy tool intended to provide a short-term economic stimulus to boost housing demand and private housing construction, it offered an interest free 'equity loan' of up to 20% of a newly constructed property's value to both first-time buyers and movers able to provide a 5% deposit. It also initially applied a generous upper price threshold of £400,000, although as noted in chapter 1, this was subsequently reduced.

Housing market developments since 2015

ONS data indicates that between January 2015 and October 2019 the average (geometric mean) nominal house price in Scotland increased by some 12% to £153,692, which was broadly in line with growth in average earnings. Real house prices also fell broadly into line with the underlying price trend since 1990.

With earnings growth (30%) having outstripped nominal house price growth (20%) in the 10 years to 2018, house prices have become more affordable. This has been most pronounced at the entry level, where price-to-earnings ratios fell from 5.7 in 2007 to 4.4 in 2018. Mortgage burdens have also fallen back. In 2015 first-time buyers spent an average of 19% of their gross income servicing a mortgage. By 2019, this proportion had fallen to 17%.

Access to owner occupation has also been improved by the relaxation of mortgage lending criteria and in particular deposit requirements. UK Finance report that during 2015-2019 the average LTV ratio for first time buyers in Scotland increased from 78% to 83%. More significant still, the Financial Conduct Authority (FCA) report that over 56% of advances to first-time buyers in the first six months of 2019 had an LTV of 85% or above, including 28% that had an LTV of 90% or above. (see figure 14).

Source: FCA Product sales data - 2019 H1. 2019 based on date for first 6 months only.

In summary, recent improvements in the affordability and accessibility of owner occupation, especially for households with a gross income of £39,000 or above (UK Finance) has been accompanied by a growth in the numbers of first-time buyers and some increase in the proportions of younger households that have become homeowners (see figure 8). FCA (2019) report that in 2018 some 93% of first-time buyers purchased a property with a value of £250,000 or less, including 42% who purchased a property valued at no more than £120,000.^[7]

Housing market activity has increased, with first-time buyers now accounting for around half of all those buying a home intended for owner-occupation. Private housing construction has also expanded, albeit numbers remain rather lower than at the start of the 21st Century (see figure 12).

Position of shared equity products within the market

An important consideration is the extent to which the shared equity products have contributed to improvements in the housing market in recent years, including developer sales. This will be explored further in chapter 8 but first it is necessary to consider how shared equity sales are positioned in the housing market.

Table 4: Completed purchases under each of the three shared equity schemes

	Private new build sales	Resales	HtB	NSSE	OMSE	All sales	HtB as % of new build	All shared equity as % of all sales
2007-08	18,537	130,606		527	654	154,403		0.8
2008-09	11,490	75,046		651	512	102,154		1.1
2009-10	8,546	63,317		721	1459	69,032		3.2
2010-11	7,755	64,288		576	579	73,875		1.6
2011-12	7,585	62,920		566	186	69,375		1.1
2012-13	7,998	65,003		375	533	72,320		1.3
2013-14	8,808	78,584	750	316	1051	84,101	8.5	2.5
2014-15	9,835	82,668	3560	252	1030	93,544	36.2	5.2
2015-16	10,894	88,392	3,690	189	1456	96,630	33.9	5.5
2016-17	11,279	88,591	2,370	177	1653	99,365	21.0	4.2
2017-18	11,895	90,293	2,290	165	1766	103,350	19.3	4.1
2018-19*	12,205	89,423	2,370	175	1797	100,998	19.4	4.3
07/08 to 12/13	61,911	461,180		3,416	3,923	541,159	-	1.4
13/14 to 18/19	64,916	517,951	15,030	1,274	8,753	577,988	23.2	4.3

Sources: Scottish Government administration records (HtB, OMSE & NSSE) and Registers of Scotland (RoS) 2019 Property Market Report, 2018-19.

One important consideration is the proportion of house sales comprised of sales assisted through one of the shared equity products, and the contribution these have made to the volume of sales achieved by private developers. Table 4 reports on shared equity sales volumes and how these compare with property transactions for Scotland as a whole. Likewise, figure 15 illustrates trends in new housing outputs and private new build transactions relative to HtB sales.

There were over 15,000 HtB sales between 2013-14 and 2018-19. Around 23% of all private new build transactions since the start of 2013/14 have been supported by HtB but this share has fallen over time. In 2014-15 HtB accounted for 36% of private new build transactions but this share had fallen to 19% by 2017-18 and 2018-19. This decline is associated with the reduction on the cap on the eligible house price from £400,000 in October 2014 to £200,000 in April 2017 plus the reduction in the upper limit in the Scottish Government's equity stake from 20% to 15% from April 2016.

Between 2012-13 and 2013-2014 the annual numbers of OMSE sales almost doubled from 533 to 1,051. Although they increased further to 1,797 sales in 2018-19, OMSE accounted for only 1-2% of second-hand sales in each of the six years to 2018-19 inclusive.

In the wake of the introduction of HtB, annual numbers of NSSE sales have fallen back. In the 3 years to the end of 2018/19, NSSE had fallen to an average of 172 each year.

Sources: Scottish Government (2019) housing statistics, and shared equity administrative data plus ROS 2019 Property Market Report, 2018-19.

Our main interest is in HtB as a share of new build transactions and the extent to which HtB has boosted developer sales but in the interests of transparency figure 15 also includes private housing construction figures. It shows that Registers of Scotland (RoS) new build transaction volumes are typically around 20% lower than private sector completions and that the scale of this gap has varied over time. One reason for this difference is that private sector led completions include some units that are built by the private sector but are not intended for sale to private households, such as self-build and properties transferred to social landlords. Another reason is that the RoS method for identifying new build properties is centred on new homes built by 'major builders' and undercounts small and single unit developments by small-scale builders. However, these two factors do not appear to fully account for the difference between the two datasets.

The launch of HtB and the expansion of the OMSE scheme in 2013-14 occurred when there were already signs of some improvement in new build transactions and private sector led housing starts and completions. The upswing in the sale of newly constructed properties has also continued despite the decline in HtB transactions. This highlights some of the challenges in trying to separate out the effects of HtB on the supply responsiveness of private developers from wider developments in the economy and the housing market.

Table 5: Average prices for shared equity and other mortgage funded sales, 2007-08 to 2018-19

	Average all property price	Average new build price	FTB avg price	Existing owner avg price	Help to Buy	NSSE	OMSE
2007-08	£157,296	£208,000	£110,227	£186,271
2008-09	£173,301	£225,000	£118,703	£201,350
2009-10	£167,736	£196,000	£121,357	£195,251
2010-11	£184,956	£200,000	£129,341	£210,648
2011-12	£181,846	£215,000	£125,291	£209,621
2012-13	£179,057	£220,000	£122,453	£209,164
2013-14	£178,109	£213,000	£128,124	£211,413	£188,000
2014-15	£189,681	£216,000	£135,631	£221,759	£203,000
2015-16	£181,220	£211,000	£141,166	£213,780	£185,940
2016-17	£183,883	£244,000	£137,255	£231,433	£179,400
2017-18	£182,492	£255,000	£135,736	£232,083	£170,300
2018-19*	£188,589	£253,000	£145,799	£234,215	£174,900
Survey (mean)					£183,860	£133,920	£120,362

Source: Scottish Government administrative data, ONS simple average sales prices, which are derived from the Regulated Mortgage Survey (RMS), buyers survey.

In terms of how shared equity products are positioned in the market, it is also helpful to look at the price of properties purchased through shared equity. Table 5 compares simple average prices of properties purchased with a mortgage with simple average prices for HtB derived from administrative records. As similar data for OMSE and NSSE prices are not available, illustrative price estimates have been derived from survey results.

The use of survey data prevents the estimation of year on year price movements, but it supports perceptions that OMSE and NSSE initiatives have, as intended, been targeted at modest income households that struggle to afford owner occupation. Table 5 shows that the average price paid by NSSE buyers (£133,920) has been below the average price paid by all first time buyers since 2014-15. Likewise, the average price paid by OMSE buyers has been below the average price paid by all first time buyers since 2009-10.

Prior to 2015-16 the average HtB sale price was above the average price for all mortgage backed sales in Scotland. However, the imposition of successive lower HtB price caps has had the desired effect of reducing average HtB prices below the average house price for all sales in Scotland. The lowering of the HtB price caps has also increased the price gap between the average HtB price and the average price for all mortgage backed new build sales. In 2013-14 the average HtB price equated to 88% of the average price for all mortgage backed new build sales but by 2018-19 this proportion had fallen to 69%. This reflects the fact that increasing proportions of new build sales sold for up to £230,000 since 2016/17 and £200,000 since 2017/18 have been assisted HtB purchases (see figure 16).

On the other hand, the average HtB price has continued to exceed the price of mortgage backed first time buyer sales. In 2018-19 HtB prices remained 20% higher than the average house price paid by all first-time buyers.

Sources: ROS house sales transactions and Scottish Government Help to Buy administrative data.

One of the rationales for introducing HtB was to help restore consumer confidence in the housing market. From a household perspective this means that HtB was intended to assist prospective buyers overcome concerns about the affordability of house prices and their ability to sustain high mortgage payments or to overcome access barriers associated with raising the necessary deposit, or a combination of both.

Price comparisons suggest that substantial numbers of HtB buyers may have been in a financial position to afford to buy without assistance. For these households the key benefit from accessing HtB may have been to remove access barriers brought about by the higher mortgage deposit requirements, which as discussed earlier, were imposed in the wake of the financial crisis. As such one of the issues the survey sought to explore was the extent to which HtB assisted households to bring forward their purchase decisions.

The Scottish Government's Help to Buy monitoring report for 2018-19 reports that:

"the mean income of purchaser households has dropped from £50,000 in 2014/15 to £42,000 in 2015/16 and then £41,000 in 2016/17, after which average incomes have remained at similar levels across 2017/18 and 2018/19. This is broadly consistent with the timing of the lowering of the purchase price cap from £400,000 to £250,000 in October 2014, and then to £200,000 in April 2016, given that there can typically be a gap of some months between the issuing of an authority to proceed with a purchase and the entry date following the sale".

Scottish Government (2019) page 5 (study team emphasis)

At £41,000, the average income is close to the average for all first-time buyers in Scotland which stood at £39,700 in 2018. Assuming a household with a gross income of £41,000 purchased a home valued at £174,000, secured a 15% equity loan and put down a 5% deposit, we estimate that a 30-year mortgage with a 2.3% interest rate would result in a monthly mortgage cost of £615. This implies that the share of gross household income spent on mortgage repayments would be 18.9%.

This lends further support to suggestions that any additionality from a demand perspective is likely to have arisen from enabling households to overcome the access barriers imposed by the imposition of strict lending criteria as opposed to affordability of house prices per se. This is an issue to which we return in chapter 8.