A Reconnaissance Study on the

Market Impacts on Elevated Homes in Known Floodplains

 

City of Snoqualmie Case Study

 

 

 

 

 

By

Ron Throupe Ph.D.

Assoc. Director

Runstad Center for Real Estate Studies

University of Washington

 

&

 

Bob Freitag CFM, Director

Institute for Hazards Mitigation University Washington

 

&

 

 Rhonda Montgomery, CFM

City of Snoqualmie

 

 

 

ã

Draft, June 27, 2002

 

Do not quote without permission of authors

 

 

A Reconnaissance Study on the

Market Impacts on Elevated Homes in Known Floodplains

City of Snoqualmie Case Study

 

By Ron Throupe, Bob Freitag, Rhonda Montgomery,

 

 

Summary of Conclusions:

 

Elevating a home above the base flood elevation is required by the National Flood Insurance Program (NFIP) for new and substantially improved structures.  Elevation through retrofitting is a common practice and has received support through the Small Business Administration, the Hazard Mitigation Grant Program and the Flood Mitigation Program.

 

This study addresses the question of what effect elevating a home has on the selling price and time on market when compared to that of non-elevated homes within the same local market. The study scope is a single incorporated area, the City of Snoqualmie, Washington. The regional housing market condition, attitudes of buyers, sellers and real estate agents were peripherally included in the study and antidotes.  Within these limitations we conclude from the community studied:

 

 

This 25% to 70 % of the retrofit costs represents a meaningful return to the homeowner when compared to an initial investment of 12% to 25% and a greater amount than would be expected if the price differential were solely a result from a reduced cost of flood insurance. 

 

The results suggest that when elevated homes are not perceived as being out of character with the neighborhood, non-elevated flood prone homes are discounted and the cost of the retrofitting approaches the value of the discount from expected market prices. 

 

There is an implication that over time there is some financial gain to the property owner, this gain does not however equal the costs involved in elevating a home.  Without state involvement there would not be the financial support to encourage significant numbers of homeowners to elevate flood prone homes.  The gains to the community from the reduced need for disaster response and recovery assistance out weigh the individual benefits to the homeowner.

 

 

SNOQUALMIE BACKGROUND:

 

The City of Snoqualmie started as a mill town at the turn of the 19th Century. This history provides much of the City’s character and identity today. The town center is a blend of museums, a historic railroad yard, a wooden monument to ancient old growth forests and a dozen or so commercial structures.

 

The original or “historic” City of Snoqualmie has a population of about 1,700 residents.  All of historic Snoqualmie except two homes are in the special flood hazard area.   This amounts to almost 700 dwelling units.  The housing reflects the modest styles that were capable of being built by the mill workers themselves – simple cottages and bungalows constructed with sturdy old growth timbers. A few post war ranches have replaced the more poorly built and flood damaged homes.

 

This once remote community continues to become a bedroom community of Seattle.  Snoqualmie was rediscovered as commuters found the 30-minute trip from Seattle increasingly manageable with infrastructure buildup, and the mill town more affordable than the closer in suburbs. An example is the annexing by the City of a newer development some years ago.  This new community is located on a ridge some distance from the original town site and is located out of the special flood hazard area.  This “Ridge” community has a totally different character and was filtered with the flood vs. non floodplain variable, Flood.  

 

The city is located on the higher foothills of the Cascade Range. Above the original town site two River forks merge with the Snoqualmie River.  The Snoquamie River forms the eastern most boundary of the city limits.  At the lower end of the city limits the River reaches Snoqualmie Falls.  Snoqualmie Falls creates a restriction that impedes the flood flows causing a backing up of the water into the town. Floodwater is relatively free of sediment, and neither its rise nor its current is life threatening. About every five years the town is flooded with possible depths exceeding six feet above grade in some housing areas. These occurrences led to the City of Snoqualmie’s inclusion in 13 presidential flood related declarations between 1965 and 2001.  

 

Retrofitting History

 

To make Snoqualmie a more livable community, many residents have elevated their existing homes.  Home elevation is a complicated process because the City is in a seismic ICBO zone 3, soon to be a designated a zone 4.  Combined with the high cost of construction in the Puget Sound area relative to other areas of the state and region.

 

About 60 homes have been retrofitted through elevation between 1987 and the spring of 2002.  Another 100 homes have been constructed since the City entered the National Flood Insurance Program requiring the homes to be elevated as part of the initial construction. The first significant elevation/retrofitting project occurred in 1987 after a Federal disaster declaration.  The City revised its Flood Hazard Regulations to define “Substantial Improvement” as any repair, reconstruction, or improvement of a structure, other than a flood-damaged residential structure, the cost of which equals or exceeds 50% of the market value of the structure.  With respect to a flood-damaged residential structure, “substantial improvement” means any repair, reconstruction, or improvement of a structure, the cost of which equals or exceeds 10% of the market value of the structure.  This definition far exceeds the minimum set by the National Flood Insurance Program.  Accordingly, the Small Business Administration made loans available to homes that met existing codes, and about a dozen homes were elevated. Another 50 homes where elevated through the Hazard Mitigation Grant Program (HMGP) after presidential disasters in the November 1995, February 1996 and Winter Storms of 1997 flooding events.

 

The HMGP program was both a federal and state sponsored program that contributed 75% - 87.5% of funds to the cost of elevating the home.  The homeowner contributed the remaining 12.5% - 25% cost of elevating depending on federal guidelines in place at the time of the award.  The City also received Flood Mitigation Assistance funding to elevate three repetitive loss homes, requiring the homeowner to contribute 25% of the costs of elevation.  In addition, six to seven elevation retrofits occurred without outside assistance with the homeowner paying the entire cost of elevating the home.

 

 

Discussion:

 

Preliminary research uncovered numerous questions to address.   The following is a sample of questions: 

 

 

Value Added

 

Retrofitting a floodprone home through elevation presents some design challenges.  For a home elevated a few feet above grade, the peculiar aspects of the elevation can be camouflaged through decking or landscaping. Elevating 4 to 6 feet is more difficult to aesthetically soften.  In Snoqualmie a combination of landscaping and decking was the common approach used to visually distract the impacts of elevation.   Elevating eight feet was impossible to conceal, but along with the challenges there were opportunities. Often the first floodprone floor could be designed to look like a lower floor and most importantly value could be added to the elevation. A garage under the “first floor” could be incorporated into the design along with additional storage space. In a couple of rare instances actual additional floor space could be added.  These were mostly  “slab on grade” homes with attached garages where the entire home was elevated including the space allocated for the garage.  A floor was constructed under the entire structure and the garage was placed under the floor.  Value added concerns were addressed to the extent possible. However, because this study was of a single community, with similar flooding characteristics and housing stock, conclusions were cautiously interpreted. 

 

Perception Through Time

 

From discussions and involvement with elevation projects over the years, it was our belief that initial elevation of a home has a neutral affect on the market price of that home. But it may raise a red flag by signaling to prospective buyers that this house and neighborhood, floods. In addition, if not well planned, the elevated house can look strange and out of character with the neighborhood and priced accordingly by the market, thus a longer time on market before a buyer is found.

 

However, as additional homes are elevated, attention is then diverted to non-elevated homes, a change in perception occurs and non-elevated homes become a concern to buyers.   At this point a threshold is reached, elevation begins to have a diminished affect on the sales price and time on market.  Then non-elevated homes become apparent and values for these homes incorporate deferred elevation remedies.

 

The time until the threshold was reached could not be answered without studying similar floodprone communities of elevated homes within the contexts of comparable non-floodprone neighborhoods.  A formal survey interview to determine attitudes of buyers, sellers, real estate agents and appraisers is planned.  Informal discussions with local realtorsã disclosed that they believe elevation of the home is being priced in the market. However, results of this investigation for one market suggest there is an evolution of acceptance.  With only 10% of the homes within the floodplain elevated, it appears that a threshold has been reached and elevated homes reflect comparable homes in comparable non-flood prone neighborhoods.

 

Return on Investment

 

Research suggests that homes located within a regulatory flood zone sell, on average, for less than homes located outside flood zones, and that the price differential is a function of the present value of future flood insurance premiums (Harrison, Smersh, Schwartz, 2001). Following this logic, an older flood prone home in Snoqualmie[1] would sell for less than a home outside of the floodplain where flood insurance would not be required. The reduction in sales price of the flood prone home would approach the cost of what the premium fees could buy in additional debt.  In Snoqualmie, an average policy for an at grade flood prone home would cost $800. This $800 represents the present value of the avoidance of flood risk[2]. Therefore, in a rational market controlling for location and amenities, one would expect such a Snoqualmie home located in a floodplain to be $9,500 less than a comparable home located out of the flood zone.  A rational home buyer, all things being equal of course, willing to buy a $200,000 home outside of the floodplain would be willing to pay only $190,500 for a home in the Snoqualmie flood zone, factoring in the added cost of flood insurance. The above mentioned research has indicated that this differential is often considerably less than our example however.

 

When we extend this logic to retrofitting we would similarly assume that retrofitting a home by elevating the first floor above the floodplain would result in a similar methodology.  In Snoqualmie, the average elevated home is several feet above the base flood elevation (BFE) and the flood insurance premium is accordingly less.  Elevation is typically not the result of fill and the home remains in a flood zone but with the first floor several feet above the BFE. The average yearly flood insurance premium for an elevated home runs about $300 per year.  

 

This present value of the $300 insurance differential is $3,600 and accordingly we can assume that the rational homeowner would be willing to elevate the first floor if doesn’t cost more than the insurance cost differential ($9,500 - $3,600) equaling $5,900. We also expect a rational buyer to pay about $3,600 less for a elevated flood prone home than a similar non-flood prone home.

 

The first retrofits were initiated in 1986 within the City of Snoqualmie. The costs ranged from $5,000 to $15,000 in 1986 dollars depending on how much of the work was done by the owner.  These homes were predominantly financed though loans provided by the Small Business Administration. Today elevation costs from $30,000 to $40,000.

Our investigation concludes that the dollar value of home elevation after subtracting for value added is proportional to the increase in market value and the increase is between 25% and 70% of the cost of elevation in today’s dollars.  Specifically, if a home cost $35,000 in today’s dollars to elevate, the home would return $9,000 to $24,000 dollars depending on house value.  On average an elevated home sold for $194,167, with an Elevation coefficient of 7.65%, it equates to a value of $14,854 as the sales price discount for not being elevated. Most of the homeowners elevated their homes through HMGP funds and were required to pay between 12.5 and 25 percent of the cost of elevation. Clearly this study supports that elevating was a sound investment by the property owner.

 

The return of 25% to 70% is in direct contrast with our expected neutral affect or said otherwise as the expected rate of increase based on the reduced cost in flood insurance.  

 

Policy Implications

 

The State of Washington is prevented from providing public investment for private gain.  Their participation in the elevation of homes is undertaken because such investment is deemed for the public good, not private gains.

 

The state and federal agencies that support flood risk reduction can learn from this investigation in several ways. 

 

1.      They can be confident that the moneys spent on elevation created usable safe dwellings.

2.      The sales price gain from elevation is within the cost of construction. 

3.      The betterment issue is not an overriding concern.

4.      There are ways to blend elevation with the home and homeowners can be informed on how to accomplish quality aesthetics.

5.      The return on the investment of the homeowner is minimal if non existent initially, and that only over time is their a realized benefit

6.      Public investment can begin a revitalization and risk reduction process that may become sustaining.

 

This study has policy implications for programs involved in flood risk reduction and neighborhood revitalization.  Among these are a review of the betterment concerns and program language as far as responsibility for costs if homes are raised above the minimum requirements of flood level and contributory gains. In addition, coordination of federal and state programs to expedite elevation.  Although this has been a successful program in Snoqualmie, less than 10% of the homes effected have been elevated and associated reasons warrant investigation.

 

 

DATa

 

In order to study the housing market conditions in Snoqualmie recent housing sales were researched.  Sales of home between 1998 and present were extracted from the Northwest Multiple Listing Service.  A follow up review was performed by physically driving each of the streets in Snoqualmie to verify the home locations and condition.  The data base of sold homes was then cross checked against the city of Snoqualmie list of elevated homes to determine how many elevated homes had been sold since it was elevated.

 

The original sample size for this study included 240 home sales from 1998 to May 2002.  The physical attributes of these homes were used to create a database that included sales price, time on market, and other pertinent seasonal and time variables.  These variables are shown in table I. The physical variables included the number of baths, bedrooms, fireplaces and whether the home had a basement or garage. Quality variables were the type of roof and siding.  A seasonal variable for which quarter of the year the home sold was implemented. Since certain seasons tend to have greater transaction volume and may have an effect on the sales price of a home sold.  The sales price, listing price, and time on market (TOM) were used to reflect current housing market conditions.  Last, a set of dummy variables was included for each year of sales.  They were used to isolate the price increase trend that was in the market, a reflection of the greater Puget Sound market which had double digit price increases per year during the time period of the home sales in this study.

 

The original sample of home sales sample was then paired down.   Several sales were removed because of missing data due to incomplete county record (3).  The sample was further reduced by excluding homes that were not in the flood prone area of the city. This eliminated the newly created ridge community that has no flooding threat.  The final study sample was 128 homes in the flood prone area of town. From this final sample we identified 15 homes which were previously retrofitted by elevating.  Table II shows the summary statistics for the study sample.

 

The average time on market (TOM) for the elevated homes was 58 days. Table II & III show the subset of homes, elevated vs. non-elevated in the flood plain with a TOM of 62.5 and 58.5 respectively.  The average selling price of elevated homes was $194,0167 and of non-elevated homes was $233,275. 

 

 

Methodology:

 

In order to further investigate the market pricing effect of elevating homes a regression analysis was developed.  A hedonic pricing model was used to investigate whether the elevation of a home was significant in the sales price of a home in the study sample.  The concept of a hedonic pricing model is well-documented in housing research literature.  The theory is that the selling price of a home is based on the components of the home, the market conditions at the time of the sale and the individual party’s personal circumstances at the time of sale. This allows the researcher to control for differences in housing.   The selling price of the home is the dependant variable and a function of explanatory variables that include physical and market attributes.  The basic equation is as follows:

 

SP =f ( H,M,I )

 

The selling price is a function of housing characteristics (H), market conditions (M) and individual participant motivations (I).

 

This study did not evaluate individual buyer and seller response; thus (I) variables are not included. 

 

SP = C + B1M­1+ B2H2 + B3 T1 + B4T2 +  e

 

Where C is the constant and represent unidentified explanatory variables.

 

B is the coefficient for each attribute

 

M is a vector of market condition variables 

 

H is a vector of market condition variables 

 

T is a set of dummy variables for years

 

e  is an error term

 

            and

 

SP is the actual selling price of the home

 

These Variables are listed below in table I

 

 

TABLE 1

 

Variable

Description

LNSP

Log of the selling price

Tom

Time on Market

Elevation

Yes =1 no = 0

Heat source

Yes = 1 for gas, 0 otherwise

Bath

Number of bathrooms

Bedroom

Number of bathrooms

Basement

Yes =1 no = 0

Garage

Yes =1 no = 0

Exterior Finish

Yes = 1 for metal or siding, 0 otherwise

Fireplace

Yes =1 no = 0

Roof

Yes =1 no = 0

Flood

1= in the flood plain, not in flood plain =0

Summer

Season of Sale, Dummy variable

Spring

Season of Sale, Dummy variable

Fall

Season of Sale, Dummy variable

Winter

Season of Sale, Dummy variable

1999

Year of sale dummy variable

2000

Year of sale dummy variable

2001

Year of sale dummy variable

2002

Year of sale dummy variable

 

 

 

 

The following are descriptive statistics for elevated homes that sold during the study period.  Of particular interest is the average selling price of these properties was $194,167 and the average time on market (TOM) was 62 days.

 

 

Table 2

a. elevated homes in the flood plain

 

 

In comparison, for homes in the flood plain, but not elevated are shown below.  The TOM for the non-elevated group was 58.5 days and the average selling price was $233,275.  Thus elevated homes surprisingly took a greater time to sell.

 

 

Table 3

a. non-elevated homes in the flood plain

 

The regression results in table 4 show the variables of significance in predicting the selling price of the house.  TOM was omitted because the literature has demonstrated that TOM and the selling price (SP) are linked with a simultaneity problem.  TOM effects selling price and  SP effects TOM.  The interpretation is that the coefficients represent percentages of the sales price. For the variable of interest, Elevation has a coefficient of –0.07636 or-7.636% of price.  The average selling price for elevated homes was $194,167, multiplied by 7.636% is equivalent to a discount of $14,815 for not elevating a home.  These results support the conclusion that elevation does pay a dividend to the owner.


Table 4

 

Regression Coefficients

 

Model variables

Coefficient

t

Significance

Constant

11.4

157.87

.000 ***

 

 

 

 

Elevation

-0.07636

-1.66

.064 *

Bedrooms

0.0851

3.56

.000 ***

Bath

.263

9.55

.000 ***

Flood

-0.0129

-.443

.658

Heat source

-0.04428

-1.096

.274

Exterior Finish

-.128

-3.767

.000 ***

Basement

0.05023

1.224

.222

Garage

-0.0384

-.997

.320

Fireplace

.123

4.147

.000 ***

Roof

.495

10.86

.000 ***

Winter

-0.0384

-.958

.339

Spring

0.055

1.456

.147

Summer

0.0247

.665

.513

Year 2000

.13

3.725

.000 ***

Year 2001

.157

3.876

.000 ***

Year 2002

.120

1.75

.001 ***

a. Dependant variable =LNSP

*** 99% level

* 90% level   

 

 

Conclusions and Further Study

 

This study was motivated by a desire to answer the question of whether the National Flood Insurance Program and the city of Snoqualmie were successful in remediating flood prone homes.  The results are evidence that the program was successful.  Buyers in the market for housing positively perceive elevated homes. The results of the statistical model shows the elevation was statistically significant and equates to a value of $14,854 for the average sales price in the market. 

 

There is a need for further study to gain additional insight into the results of the national flood program and home prices.  An expanded study of multiple markets is desirable. This expanded study could investigate whether the pricing of elevated homes is market specific or a larger phenomenon. The use of statistical methods would have greater support because a larger sample size would lend itself to investigating time specific periods and the evolution of the acceptance of elevation.

 

A formal survey of previous buyers and sellers would reveal the motivations and knowledge of these groups.  The survey would also give further insight into whether parties in a transaction were fully informed of flooding potential in the market and if they believe they priced that risk in their transaction price.  This transaction pricing questioning can be approached using contingent valuation (CV) methodology. This methodology is commonly used in natural resource studies.  It questions the individuals  ‘willingness to pay” to avoid a particular circumstance such as polluted air, water, or in our case flooding.   These approaches will help answer the following.

 

Future study questions:

 

References

 

Felton, R. S., W. K. Ghee and J. E. Stinton, A Mid-1970 Report on the National Flood Insurance Program, Journal of Risk and Insurance, 1974, 41:4, 579-99.

 

Harrison, David, Greg Smersh, and Arthur L. Schwartz. Environmental Determinants of Housing Prices: The Impact of Flood Zone Status, Journal of Real Estate Research, 2001, 21:2, 2-20.

 

MacDonald, D.N., H. L. White, P. M. Taube and W. L.Huth, Flood Hazard pricing and Insurance Premium Differentials: Evidence From the Housing Market, Journal of Risk and Insurance, 1990, 57:4, 654-63.

 

Powers, F. B. and E. W. Shows, A Status Report on the National Flood Insurance Program Mid 1978, Journal of Risk and Insurance, 1979, 46:2, 61-76.

 

Pritchett, S.T. and H. W. Rubin, A Case Study of Flood Losses: Implications for Flood Insurance Product Development, Journal of Risk and Insurance, 1975, 42:1, 105-115.

 

Shilling, J. D., J. D. Benjamin and C. F. Sirmans, Adjusting Comparable for Floodplain Location, Appraisal Journal, 1985, July, 429 –36.

 

Shilling, J. D., C. F. Sirmans and J. D. Benjamin, Flood Insurance, Wealth Redistribution, and Urban Property Values, Journal of Urban Economics, 1989, 26, 43-53.

 

Skrantz,T.R. and T, H. Strickland, House Prices and Flood Event: An Empirical Investigation of Market Efficiency, Journal of Real Estate Research, 1987, 2:2, 75-83.

 


[1] Pre FIRM: Homes build before the adoption of the Snoqualmie Flood Insurance Rate Map.  

[2] Estimate of a 7% interest rate over 30 years.