Special Studies, January 9, 2009
By Paul Emrath, Ph.D.
Report available to the public as a courtesy of HousingEconomics.com
Even when nothing else changes, a policy that effectively increases the number of home owners or otherwise stimulates home buying activity in the U.S. will impact housing markets and the overall economy in a number of ways. Some of the housing units currently being rented will become owner-occupied. Vacant units for sale on the market will be absorbed more quickly. Because a share of first-time buyers typically chooses to buy new homes, construction will be stimulated and directed toward particular types of housing. As owners of existing homes are displaced, it will set up a “ripple” effect of trade-up buying. The new construction generated in both the initial and ripple phases will in turn generate income and employment in various industries, as well as tax revenue for federal, state, and local governments. The home buying activity will generate fees for real estate agents, income for the professionals who provide services at closing, and additional spending on items such as home appliances and furnishings.
This article describes the techniques NAHB uses to estimate these effects. To illustrate the techniques, the article shows the estimated economic effects of a recently enacted policy designed to increase homeownership - the homebuyer tax credit created by the Housing and Economic Recovery Act of 2008. The credit established by this Act is for $7,500 (or 10% of the price, whichever is less) and remains in effect through June of 2009. Because the credit must eventually be repaid, even if over an extended period of time, it is more of an interest-free loan than a tax credit in the conventional sense. This “credit” is restricted to first-time homebuyers; so, to the extent the credit achieves its objective, it will increase the number of homeowners in the U.S. A previous article estimated that the first-time homebuyer credit will result in 143,000 new home owners. The 143,000 is not an estimate of the number of buyers who will claim the credit on their tax forms, but of the households who will be motivated to buy their first homes only because the new credit has become available.
Assuming that the estimate of 143,000 new first-time buyers is correct, the additional estimated first-year effects of the new homebuyer tax credit include the following:
- A ripple effect of 83,000 additional home sales and acquisitions generated by displaced home owners trading up.
33,000 of the vacant units currently on the market sold and occupied by home owners.
25,000 renter-occupied units (many probably being rented because owners are currently unable to sell) sold and occupied by home owners.
- A net increase of approximately 41,000 additional housing units built, with a net value of $7.0 billion.
An additional increase in economic activity from broker and other transaction-related services, property alterations, and furnishings of $3.0 billion.
- Creation of approximately 85,800 full-time jobs (enough work to keep one person employed for a full year), $4.2 billion in wages and salaries, $2.0 billion in small business income, and $1.4 billion in corporate profits.
Additional tax revenues of $2.3 billion for the federal government (helping to off set the cost of the program to U.S. tax payers) and $700 million for state and local governments.
Subsequent sections of this article explain how these impacts are estimated and provide additional details, including a breakdown of the jobs generated by industry.
Home Sales and Acquisitions
Table 1 shows the estimated impacts of the first-time homebuyer tax credit on sales and acquisitions by type of home purchased or acquired. The term “acquired” is used to account for new custom homes. As used here, the qualifier “custom” refers to homes built on the owner’s lot, where the owner either hires a general contractor or functions as the general contractor him or herself. Technically, custom homes aren’t counted as part of new home sales, but they need to be included among the types of construction that home buying activity may stimulate.
Estimates in the first numeric column of Table 1 show that the largest share of the new first-time buyers (100,000 of the 143,000) will purchase existing single family homes, followed at a distance by purchases of new single family homes built for sale (15,000) and existing multifamily condominiums (11,000). These estimates were derived from tendencies of recent first-time buyers computed from 2007 American Housing Survey (AHS) data. The AHS is based on a nationally representative sample of housing units, funded by the Department of Housing and Urban Development, and conducted in oddnumbered years by the U.S. Census Bureau. The 2007 AHS tendencies were adjusted to account for the trends of declining homeownership and increasing vacancies that continue to persist in the weak economic and housing-market environment. The adjustments to the AHS tabulations are based on quarterly numbers published by the Census Bureau from its Housing Vacancy Survey. Compared to unadjusted estimates derived from 2007 data, the table shows a few more renter-to-owner conversions and a substantially greater number of vacant units absorbed.
First-time buyers who buy homes that were previously owner-occupied generate ripples of trade-up buying, and estimates of this effect are shown in the second numeric column of Table 1. Several timing issues arise when estimating these ripples. The numbers presented here are first-year estimates that assume the 143,000 induced sales and acquisitions by first-time buyers take place at a constant rate during the year. Hence, the ripple effect has, on average, half a year from its inception to propagate.
The rate at which ripples propagate through the housing market also helps determine the size of the first-year effects. Ripples occur because some first-time buyers displace existing homeowners, who then need to move into other housing units. A share of these displaced owners will displace other existing homeowners, who must then move into different housing, displacing yet other owners, and so on. Eventually, the ripples die out and the total number of displaced owners converges to a limit. Because the process of buying and selling homes often requires considerable time, however, it is unlikely the market will move all the way to its theoretic limit during the first year. The estimated first-year ripple effect of 83,000 home sales and acquisitions assumes that, on average, it takes 5.3 months to sell an existing home. This assumption is based on the average month’s supply reported by the National Association of REALTORS® over the period from 1999 to the present (the period for which a consistent series including both single family and condominium sales is available).
A share of the homes purchased by new first-time and ripple-effect trade-up buyers represents new construction. Table 1 shows that the first-year effects of the homebuyer tax credit include the production of 39,084 new single family homes, 2,157 new multifamily condominiums, and 2,345 new units of manufactured housing. However, the expansion of the single family stock and the purchase of new manufactured homes will result in the need to either reduce the rental housing stock in other ways or slow the net addition of rental housing units to the stock.
Two responses are likely to occur that will mitigate the reduction in rental multifamily construction. The first is a response in the rate of household formations. An expanding opportunity for home ownership tends to encourage existing households to split. Young people are able to move from their parent’s home or out of a shared housing arrangement. Families that are doubled up can each find its own place to live.
Also, some multifamily homes will cease to be useful and will be demolished or converted to non-residential use. As families move out of rental units into home ownership, the best rental units will remain occupied and the worst ones will fall out of the stock.
An estimate of the overall impact on multifamily rental production is derived from the relationship between homeownership and the number of multifamily starts intended for rent during the period from 1998 through 2004. This is a period when the homeownership was rising steadily, and the rate of multifamily production was relatively constant but tilting away from rental housing in favor of condominiums. A straight line fit through the relationship estimates that one effect of the first-time homebuyer tax credit will be a reduction of 2,312 rental multifamily housing units built. Details of the calculation are shown in Table 2.
The number of new housing units combined with their estimated value can be used to estimate the value of the construction activity that the first-time homebuyer tax credit will add to the economy during the first year. When deriving this estimate, in addition to netting out the effect of reduced multifamily rental production, it’s necessary to account for the difference between the market price and construction value of a home. The difference is the value of raw land, which is not produced as the result of any construction activity (although all costs associated with developing the land do count as construction value). The accounting is done by applying the adjustments the Census Bureau makes to house prices before feeding residential construction data into the GDP accounts. Details of these adjustments were shown in a previous article.
The result is an increase in the value of new construction (net of the reduction in multifamily rental production) of just under $7.0 billion. This represents an increase in the residential fixed investment component of GDP stimulated by the first-time homebuyer tax credit.
The value of new residential construction is only part of the estimate impact the firsttime homebuyer tax credit will have on GDP, however. The induced home purchases will result in new mortgages, services provided by real estate agents, and other spending connected with buying and financing a home. Most home purchases are financed with mortgages, and financing arrangements typically involve a 1 percent fee for the lender that offsets the cost of origination and approximately 2 percent in additional settlement fees to appraisers, surveyors, credit bureaus and other settlement agents. These percentages were supplied to NAHB by the Mortgage Bankers Association. Applying them to the home sales discussed previously produces an increase in total mortgage origination expenditures of about $307 million.
Most home sales also involve a real estate agent who is paid out of the proceeds from the sale. Assuming an average fee of 6 percent on the sale of existing homes (consistent with the treatment of these fees in GDP accounts), the added economic activity is $1,671 million.
Moreover, when families buy homes, they often spend more on home repairs and alterations compared to otherwise similar households who do not purchase homes. Even buyers of new homes sometimes add decks or porches, or otherwise spend to enhance their yard in any number of ways. Regression estimates based on data from the Bureau of Labor Statistics’ Consumer Expenditure Survey show that the amount spent on remodeling, repair and alterations depends upon whether or not a home was recently purchased, whether the home purchase is a new or existing home. These results were discussed in a previous article. They imply that the homebuying activity stimulated by the tax credit will increase spending on remodeling by $150 million the first year.
Additional spending is triggered as homebuyers spend money on furnishings and appliances when they first move into their newly purchased homes. First-time buyers often need to purchase items such as lawn mowers, tools, and appliances. Even buyers moving from one owned home to another often decide to purchase different furniture and decorations as part of the transition. Regressions similar to the ones used to estimate changes in remodeling can be used to estimate the size of these effects. The estimates show that the families induced to purchase homes by the first-time homebuyer tax credit will spend about $63 million more on appliances and $146 more on furnishings than they would had they not moved.
Additional regressions found no corresponding reduction in spending on other items among homebuyers, so (given that we are netting our statistically the amount households typically spend on various items when not purchasing a home) all of this spending on remodeling, appliances, and furnishings is a net addition to GDP.
In total, the first-time homebuyer tax credit is estimated to increase GDP by roughly $3.0 billion beyond the value of new construction - from items such as mortgage origination and real estate broker fees, services provided at closing, and spending on appliances, furnishings, and property alterations that are not included in the price of the home.
Employment and Income Impacts
Except for the amount consumed by depreciation and a statistical discrepancy, every dollar of GDP becomes income for some person or business. Hence, much of the $7.0 billion in new construction and $3.0 billion in additional spending generated by the firsttime homebuyer tax credit can be converted to components of income, such as wages and salaries, corporate profits, and income earned by owners of small businesses.
Table 4 shows these income impacts by industry, and also shows the number of full-time jobs supported by the wage and salary estimates. The jobs are expressed in “full time equivalents”, meaning enough work to keep one laborer employed full time for a year.
Summed across all industries, the estimated first-year impacts of the new homebuyer tax credit include 85,775 full-time jobs, $4,152 million in wages and salaries, $2,027 million in small business income, and $1,421 million in corporate profits. Although a large share of the jobs (38,672) are in construction, the employment effects are broadbased. Indeed, the cost of materials accounts for about 30 percent of the price of a typical home. This means that homebuilding indirectly employs the workers involved in manufacturing lumber, wallboard, windows, doors, cabinets, fixtures, and other items that go into a typical home - as well as the workers involved in harvesting and extracting the raw materials that go into these manufactured products. Table 4 shows that the first-time homebuyer tax credit will support 15,987 jobs in the manufacturing and raw materials sectors of the economy.
The tax credit also supports 11,572 jobs for people who selling, transport, and store products - a large share of these in retail trade, and 19,544 jobs for service providers - such as lenders, architects, real estate agents, lawyers, and support staff who provide services to home builders and home buyers.
These job totals exclude small mom-and-pop businesses with no payroll - i.e., people who are self-employed. In residential construction, there are many such small businesses, especially among the specialty trade contractors. In Table 4, this shows up as a large value for proprietors’ income, relative to the wages and salaries generated, in the construction industry.
The ratio of proprietors’ income to wages and salaries is even more skewed for real estate brokers and appraisers. This occurs, because there is a strong tendency in this segment of the industry for individual agents to be organized as independent contractors who are not technically counted as employees of a business. Whether subcontractors in the construction industry or individual real estate agents, none of these self-employed arrangements are counted as jobs in Table 4, although many observers would probably think of them that way.
Impacts on Government Revenue
Other economic effects of the first-time homebuyer tax credit include revenues generated for federal, state, and local governments. The wages and salaries discussed in the previous section are subject to federal and state income taxes, as well as taxes for Social Security. Profits earned by both small businesses and large corporations are similarly taxed. States often impose sales taxes on material sold to home builders, many local jurisdictions levy fees for approving building permits and extending public services, and materials such as lumber and concrete are often subject to some form of tax when imported into the U.S.
Estimated taxes and other forms of government revenue generated by the first-time homebuyer tax credit are shown in Table 5. At the federal, state and local levels combined, the first-year impacts are a total of $2,988 million. About $2,284 million of this is revenue for the federal government, which helps to offset the cost of the revenue the government loses by giving homebuyers a tax credit. Technically, Congress is not allowed to consider offsets like these when calculating the cost of a stimulus package to the American taxpayer. Nevertheless, it is often helpful to provide individual members of Congress with credible estimates of ancillary revenues when advocating in favor of polices to increase homeownership or otherwise stimulate home sales.
Tables 1-5 (Excel)
See other Special Studies