Factors driving Irelands Housing Need David Duffy

Factors driving Irelands Housing Need David Duffy

Residential Property Research Programme 2 Year Research Programme Funded by NAMA and IBF ESRI undertake research and publish findings Year 1 focus on demand

Outline Forecasts for new household formation Population forecasts Headship Alternative assumptions Where will the dwellings be needed? Costs and new household formation

Population forecast ESRI Demographic model Population forecasts based on life expectancy, fertility, births, and deaths (CSO) migration, education Population in each year by single year of age and gender

Millions Population forecast Recovery Scenario, MTR 6 2025: 5.01m 2030: 5.16m 2020: 4.84m 5 4 2011: 4.6m 3

Thousands Population Structure, aged over 25 yrs 1200 1000 800 600 400 200 0 25-29 yrs 30-34 yrs 35-39 yrs 40-44 yrs 45-49 yrs 50-54 ys 55-59 yrs 60-64 yrs 65 yrs + 2011 2020 2025 2030

Headship and Household formation Headship rates measured by the proportion within each age group identified as head of household or household reference person Projected forward based on trend between Census 2002 and 2011 (UN, ONS)

1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 2019 2021 2023 2025 2027 2029 Headship shows steady rise 0.5 0.45 0.4 0.35 0.3 0.25 0.2

Number of households, 000s Scenarios for New Household Formation in Ireland 2,200 2,000 1,800 1,600 1,400 2011 2014 2017 2020 2023 2026 2029 Headship Trend Census 2002 to 2011 Constant headship

Number of households, 000s Alternative Scenarios for New Household Formation in Ireland 2,400 2,200 2,000 1,800 1,600 1,400 2011 2014 2017 2020 2023 2026 2029 Headship Trend Census 2002 to 2011 Net emigration Constant headship Net immigration

Alternative Scenarios for New Household Formation in Ireland Number of Households, 2030 Increase, 2011-2030 Annual average Increase, 2011-2030 Base: Headship constant at 2011 rates 2.01 million 356,000 19,400 Trend in headship: Census 2002-2011 2.14 million 481,000 25,600 Trend in headship: Census 2006-2011 2.29 million 635,000 33,300 Converge to 2011 headship rates in England and Wales by 2030 2.15 million 499,000 26,500 Higher Immigration, +15,000 per annum 2.26 million 607,000 31,900 Higher Emigration, +15,000 per annum 2.01 million 355,000 19,000

Residential construction Under a range of scenarios over 20,000 new households, on average, will be formed each year over the long term Even with forecast growth, the level of new house completions will be lower than demand in the short-run

House completions 100,000 90,000 80,000 70,000 60,000 50,000 40,000 30,000 20,000 10,000 0 1970 1974 1978 1982 1986 1990 1994 1998 2002 2006 2010 2014

County Level Analysis There is significant difference across the country with respect to housing market and economic performance This implies that there is no such thing as a national housing market. Analysis at the county level considered the requirement for additional housing due to population growth and modelled key housing variables (see Morgenroth, 2014).

Population Projections Constructed using ESRIs county population model (IC-POP). The future population is equal to the current population (base 2011) plus the natural increase (births-deaths) plus net migration. Population projections are calculated applying assumptions about mortality (same as CSO), fertility (unchanged from 2011), and internal migration (towards the large cities) and international migration (ESRI MTR Recovery Scenario). Applying the average county household size (assumed to be declining at the same rate based on ESRI MTR), the number of households is calculated.

Dublin Cork Kildare Galway Meath Wicklow Louth Limerick Westmeath Wexford Waterford Kilkenny Clare Donegal Laois Cavan Sligo Mayo Offaly Carlow Kerry Monaghan Longford Leitrim Roscommon Tipperary Where will the households be? 9000 8000 7000 6000 5000 4000 3000 2000 1000 0-1000

Projecting Housing Needs Using the projections it is possible to estimate the required housing stock. Census 2011 identified a significant excess in vacant dwellings (vacancies are never zero). There has been some construction. There is a normal rate of obsolescence (properties become structurally unsound or are destroyed/damaged by fire etc.). Assuming no completions from 2013, accounting for initial number of vacant units, depreciation and a normal vacancy rate, calculate the number of properties required. Bulk of requirement (60%) in Dublin and Dublin commuter belt (26%)

Household formation and tenure choice Our context is Ireland's property crash Falling rents and falling prices We model housing market behaviour of young Irish people Whether to form a household What type of tenure to choose

National House prices January 2005 =100 140 120 100 80 60 40 20 0

% change in nominal house prices 1995 Q1 2007 Q1 200 180 160 140 120 100 80 60 40 20 0 2007 Q2 2013 Q4 10 0-10 -20-30 -40-50 -60-70

Affordability: Prices and Rents 100.0 80.0 60.0 40.0 20.0 0.0 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 20072008 2009 2010 2011 2012 2013 PRTB Rent Index CSO National House Price Index Q1 2014

Household growth continued in spite of housing market boom and bust Census 2011: Number of households increase by 187,112 from Census 2006 Number of households renting increase by 152,000 Homeownership rate falls from 74.7% to 69.7%

Strategy Focus on those aged from 20 to 39 (5 yr age groups) This allows us to identify distinct behaviours from each age-group QNHS Data pooled over years 2001-11 provides large samples Tenure choice as a function of the relative costs of renting and owning Difference of user cost of housing and average rents User cost: function of price expectations, mortgage costs, taxation and depreciation

User cost of housing vs. rents, 2001-2013

Results - Formation Find expected negative relationship between affordability and formation rates If rent rises or incomes fall, forming a household becomes less affordable Highest formation rates among women, migrants, employed, those in relationships Decomposition of change in formation rates: Pre-crash: change in distribution of characteristics Post-crash: change in behaviour

Results - Tenure Identify the tenure equation using the relative cost of renting and owning Find expected positive relationship between cost difference and renting Renting more likely when the user cost rises or rents fall Equivalently, negative relationship with owning New entrants to housing market tend to be renters

Conclusions Household growth continued in spite of housing market boom and bust Demographics alone cause 20,000 households to form per year Incentives favoured renting rather than owning post-crash Cost and perceptions Irish young people respond to incentives, switch to renting Fall in homeownership rate partly attributable to household formation Entrants to the housing market tend to be renters Rents have begun to rise since 2012 Household formation rate has fallen slightly in 2013 Price expectations have reversed Underpin demand for owner-occupied housing

Factors driving Irelands Housing Need David Duffy