Executive Summary
The United States residential construction industry, traditionally the largest end-use market for softwood lumber, has been undergoing a period of change for more than a decade. Builders' acceptance of substitute materials and new innovations has increased, providing a unique challenge to softwood lumber producers. In such a situation, understanding the ways in which residential builders specify and use softwood lumber and lumber substitutes is essential to the success of any softwood lumber manufacturer. The Center for International Trade in Forest Products (CINTRAFOR) completed its first study on material substitution in 1995 (CINTRAFOR Working Paper No. 57), providing a benchmark for softwood lumber use in structural applications in residential construction.
This was followed by surveys looking at material use in 1998 (CINTRAFOR Working Paper No. 73) and 2001 (CINTRAFOR Working Paper No. 93). The current study, undertaken in 2005, represents the fourth in this series and is intended to describe the trends in material use and substitution in the residential construction industry in 2004. The 2005 survey also explores builders' awareness, usage and perceptions of certified softwood lumber and sets a baseline for tracking certified lumber usage in residential construction.
In 2004, single family construction accounted for over three-quarters of construction firms' revenue. The larger firms reported a high proportion of new single family housing than their smaller counterparts. Additionally, there appears to be a negative relationship between firm size and the amount of revenue generated from repair and remodel activities. These findings are consistent with previous survey findings. The 2005 survey also reveals that the larger firms are more involved in non-residential construction. Builders in the southwest region of the US reported significantly less involvement in the repair and remodeling sector and significantly more involvement in the non-residential sector. The share of single family construction reported by builders was found to be consistent across all regions.
A longitudinal analysis (from 1998 to 2004) of substitute material usage revealed that the largest changes occurred in the usage of glulam beams, LVL, steel framing, wood I-joists, open-web trusses, and structural insulated panels, with wood I-joists, LVL, steel framing and glulam beams recording significant decreases between 2001 and 2004. In contrast, structural insulated panels, panelized wall systems and open-web trusses have experienced an increase in use since 2001. Survey respondents in the southeast and southwest regions of the country reported a steady increase in their usage of finger jointed lumber between 1998-2004. However, the usage of finger jointed lumber nationally was found to be constant over this period. The usage of glulam beams decreased substantially in the eastern US while remaining fairly constant in the western US. Nationally, glulam beams recorded the largest drop in reported use (12.6%). Use of wood I-joists declined across all regions, with the exception of the northwest, as solid wood joist prices moderated. In addition, use of Parallam TM remained relatively constant between 1998 and 2004, whereas the use of TimberStrand TM lumber increased in the western US while declining in the eastern US. The use of non-wood material substitutes (steel framing and reinforced concrete) generally declined in the southern regions and increased in the northeast.
An analysis of material usage within specific end-use applications revealed that softwood lumber use has either increased or remained relatively constant in all applications with the singular exception of load bearing walls. For headers, wall framing and roof framing applications, softwood lumber remained the dominant material with a market share of more than 70% in each application. For floor framing, the market is split between softwood lumber, wood I-joists and open-web trusses. However, it should be noted that for all structural applications, softwood lumber recorded the largest market share. The market share for softwood lumber increased in floor and

roof framing applications, remained constant in header and non-load bearing wall applications and declined in load bearing wall applications. In wall framing applications, none of the substitute materials had a market share of more than 6% whereas softwood lumber (both solid sawn and finger-jointed studs) enjoyed a market share of approximately 88.3% and 80.9% in non-load bearing and load bearing wall applications, respectively. The usage of softwood lumber in floor framing increased from 39% in 2001 to 43% in 2004, making softwood lumber the primary material for floor joists. Significantly, the market share for wood I-joists in flooring applications (its major market) declined by almost 12%. The use of wood trusses for roof framing has experienced a steady increase since 1995, rising from a market share of 46% in 1995 to 53% in 2004.
Builders rated strength, straightness, lack of defects and the availability of softwood lumber as the most important attributes of softwood lumber; a result that has been consistent over the course of the four surveys. The importance ratings for two attributes, price and price stability, have begun to decline in importance. On a positive note, home builders consistently expressed higher satisfaction levels with all of the softwood lumber attributes in the 2005 survey. A review of the data shows that the respondents consistently recorded higher satisfaction levels for all the softwood lumber material attributes between 2001 and 2004. The 2005 survey also marks the first time that builders indicated satisfaction with two important softwood lumber quality attributes: lumber straightness and lack of defects. In all of the previous surveys, builders had consistently indicated dissatisfaction with both of these attributes. The fact that straightness and lack of defects are ranked as two of the most important lumber attributes, combined with the large increase in the satisfaction ratings for both of these attributes, suggests that builders have begun to view softwood lumber as a much better value over the past several years.
It appears that builders are becoming more conscious of the environment and that this is beginning to influence the types of materials specified by some builders. Unfortunately, builders are receiving mixed messages about the environmental performance of non-wood materials. The results of this survey suggest that builders perceptions of the environmental performance of non-wood materials improved slightly between 2001 and 2004 whereas it decreased substantially for wood-based structural materials. With the exception of SIP's, all of the substitute materials are considered to be more enviromentally friendly than softwood lumber. This result sugggests that it is important that the forest products industry in general, and softwood lumber manufacturers in particular, continue to educate builders about the environmental benefits of using wood relative to non-wood materials.
A new section of the 2005 survey considered home builders awareness and use of certified lumber. The results of the survey showed that only 40% of homebuilders indicated that they were aware of certified wood. On average, only about 14% of homebuilders indicated that they have used certified wood. Among the users of certified lumber, the average percentage of homes framed with certified lumber was approximately 50%. Almost 15% of the builders who have used certified wood reported that they framed all of their houses with certified lumber.
Further, in looking at certified wood awareness and use within individual states, it was noted that awareness of certified wood was much higher in the states along the west coast of the US relative to states in the central and eastern US. About 77% of builders surveyed on the west coast reported that they were aware of certified wood products. Similarly, among the builders who were aware of certified wood, the percentage of builders who actually used certified softwood lumber was also much higher for builders in the west coast states (70%) relative to builders in other parts of the country. Previous research has shown that the willingness of customers to pay higher prices for certified wood plays a major role in the usage of certified lumber. This research shows that only 17% of the respondents in the eastern states and 29% of the respondents in the central states believe that their customers would be willing to pay higher prices for homes built using certified wood products. The percentage for respondents in the west coast states was higher at 50%. These survey results suggest that the awareness and usage of environmentally certified wood among builders is much higher on the west coast relative to the rest of the country.
The survey results suggest that in the future large home builders may well lead the effort to increase the use of certified wood in building homes. This observation is based on the fact that 67% of large builders have heard of certified wood (this represents the largest segment for this question), 43% have used certified wood to build homes (this is the second largest segment for this question), 50% think that their customers would be willing to pay a premium for a home built from certified wood (this represents the largest segment for this question) and 75% expect that their use of certified wood will increase in the future (this represents the largest segment for this question). Further research is needed to understand home builders' motivation for using certified wood and to

explore the relationship between the use of certified wood and regulatory factors (such as green building codes and efforts to improve the energy efficiency of residential homes).

Executive Summary
The US residential construction industry, traditionally the largest market for softwood lumber, has undergone a period of uncharacteristically rapid change over the past decade. The effects of timber harvest restrictions in federal and state forests on softwood lumber price, price volatility, and product quality, combined with technological advances by producers of substitute materials, have contributed to increased use of material substitutes in residential construction. The objective of this research was to assess the extent of material substitution in residential construction and provide insight into the factors driving these changes. The results offer convincing evidence that softwood lumber has continued to lose market share in the residential construction industry and that builders remain concerned about its quality and price. The study also shows a shift on the part of builders towards a more favorable impression of the environmental impacts of substitute products, including steel and concrete, relative to softwood lumber.
This study is based on a random sample of 2,400 residential construction firms segmented by geographic region and firm size. The survey was also mailed to the 100 largest home builders, as reported in Builder magazine. The overall response rate was 12.8% (12.1% of the random sample and 37.1% of the 100 largest firms). The results show that residential builders have steadily increased their use of substitute structural materials since 1995. Respondents reported increased use of all of substitute materials included in the survey. Almost all respondents reported using at least one substitute material (compared to 91% in 1995) and over 80% of the respondents reported using glulam beams, wood I-joists, and laminated veneer lumber (LVL).
While use of steel, reinforced concrete and plastic-fiber lumber increased, engineered wood products emerged as the clear winners. On a regional basis, builders in the western US reported higher usage of all substitute products. In addition, the survey data suggest that large firms were more likely than small firms to try new substitute products, particularly finger jointed lumber, structural insulated panels, laminated veneer lumber, as well as newer engineered wood products such as parallel strand lumber and laminated strand lumber.
The survey data were analyzed to assess the extent to which various structural products were used in walls, floors, and roofs, the three end-use applications that consume the greatest volume of structural lumber. The most commonly used products were softwood lumber, steel lumber, finger-jointed lumber, wood trusses, LVL, and wood I-joists. While softwood lumber still dominated wall framing in 1998, with an 83% market share, it has lost market share (down from 93% in 1995), particularly among large firms. Softwood lumber’s share of the floor framing market declined from 59% in 1995 to 42% in 1998. While it is still the most widely used product, with a 42% market share, the market share of wood I-joists has increased from 23% in 1995 to 39% in 1998. Softwood lumber used to frame roof rafters is no longer the dominant material used in residential roof systems. Survey data show that wood trusses increased slightly from 46% to 48%, while softwood lumber declined from 51% to 40%.
To assess builders’ satisfaction with softwood lumber, respondents were asked to rate the level of the importance, and their corresponding level of satisfaction, with 13 softwood lumber attributes. The importance ratings obtained in 1998 were virtually identical to those reported in 1995. Softwood lumber straightness, strength, availability, and lack of defects were rated as the most important attributes. The survey data suggest that price is much more important to large firms than small firms. Builders reported that, while they were more satisfied with the price and price stability of softwood lumber in 1998 relative to 1995, they remained unhappy with softwood lumber quality, particularly with respect to lumber straightness and overall occurrence of defects.

A gap analysis highlighted the difference between the mean importance ratings (where 7 indicates “extremely important” and 1 indicates “not important at all”) and the mean satisfaction ratings for each product attribute (where 7 indicates “extremely satisfied” and 1 indicates “extremely dissatisfied”). Survey findings indicate that while builders are less concerned with price issues than in 1995, they remain very concerned about the perceived decline in softwood lumber quality. The data provide clear evidence that residential home builders are least satisfied with product attributes they rate most important, suggesting that builders are dissatisfied with the value (defined as the ratio of price/quality) of softwood lumber.
To provide a more concise interpretation of the importance and satisfaction of the different softwood lumber attributes, a factor analysis was performed to group together those softwood lumber attributes that are highly correlated to each other. The results of the factor analysis are almost identical with the results obtained from the 1995 survey and suggest that the 13 product attributes used to describe softwood lumber can be summarized into three factors: quality attributes, economic attributes, and technical attributes.
Finally, the survey assessed builders’ perceptions of the environmental impact associated with using substitute products relative to softwood lumber. Although environmental marketing is not prevalent in the US forest products industry, most industry observers believe that it will become more important. While reduced environmental impact had the lowest importance rating of the 13 softwood lumber attributes, survey findings revealed that more builders in 1998 had a favorable perception of the environmental impact of substitute products, including steel and concrete, over softwood lumber than in 1995.
This survey clearly indicates that softwood lumber has continued to be displaced by substitute materials in segments of the residential construction industry that it has traditionally dominated: walls, floors, and roofs. To a large degree, this loss of market share can be attributed to a perception among residential builders that the value of softwood lumber has declined: a direct result of rising prices and a perceived drop in lumber quality. Much of the loss in market share experienced by softwood lumber can be attributed to the increased use of engineered wood products. Many would argue that this is a normal process of product evolution within the forest products industry, attributed to technological advances in manufacturing processes driven by the changing forest resource. However, this study identified two trends that should concern managers in the forest products industry. First, the use of non-wood substitute building materials has increased significantly since 1995. Second, there is a growing perception among home builders that using non-wood building materials (including steel and reinforced concrete) is better for the environment than using softwood lumber. This trend away from wood products is likely to continue unless there is an effective response to the challenge posed by substitute materials.

Executive Summary
A sizable literature concerned with technological substitution modeling exists within the domain of forest products. These models have generally been used to develop market share forecasts for various forest products and their substitutes based on relative product prices. Substitution models usually assume that the potential market size is known and that products can freely substitute for one another. A small but growing literature concerned with the diffusion of new innovations also exists within the domain of forest products. This diffusion literature typically focuses on factors affecting consumer acceptance for product innovations and forecasting the level of demand growth without constraining the potential market size. In this paper, we examine the dynamic sales behavior of three and four successive generations of structural wood panel products using varying forms of a multigeneration diffusion model. The multigeneration diffusion model introduced here, which encompasses the elements of diffusion and substitution modeling, assumes that a new structural wood panel product will diffuse through a population of potential consumers over time and that market share competition will be introduced with successive generations of structural wood panels.
 
Estimation results indicate that market share competition between various structural wood panel products are differentially affected by substitution and diffusion effects. The model results reveal that the aggregate market share growth and decline for southern pine plywood can be attributed mostly to substitution effects (i.e., substitution between western and southern pine plywood), while the aggregate market share growth of oriented strandboard can be attributed to diffusion effects.
 
The model results also suggest that structural wood panel products act as complements rather than as substitutes to one another. Caution should be used, however, in interpreting these results since we evaluate the structural wood panel market in the aggregate rather than evaluating specific end-use markets.
Nevertheless, market aggregate complementarity has been found in other research examining the market share competition between structural wood panel products.
 
In the near-term, the multigeneration diffusion model suggests that the southern pine plywood market has reached its peak production level over the past five years, with production forecast to decline slowly but steadily over the next decade. Western plywood is forecast to continue its downward production and market share trand. Oriented strandboard is expected to remain entrenched in a growth phase over the next five to ten years.
 
We explore several managerial implications of the model results and suggest alternative multigeneration diffusion models that could be developed for structural wood panel products.
 
Keywords: adoption, diffusion, substitution, market share competition, forecasting, technological progress, industry evolution, product life cycle, plywood, oriented strandboard, waferboard
 

Abstract
Wood product markets are subject to substitution by lower cost non-wood materials. Alternative materials such as aluminum and vinyl are generally more fossil fuel intensive and have different environmental impacts from renewable wood resources. Increased attention to environmental impacts has drawn increased attention to substitution of wood by non-wood products.
 
This study surveys the available research on explanations for substitution related to economic and technological change, then models substitution within a particular end-product market. The question of whether substitution can be explained by prices of technological change almost independent of prices is an important factor in analyzing the impacts of policies on the environment.
 
This report uses as a case study the United States residential window market, where wood and aluminum windows have been competing for the past five decades. Prior to World War II, wood windows dominated the window market. Since then aluminum windows have increased market share. By 1982, wood windows held only 26 percent of the market. Throughout much of this period, not only were wood windows more expensive than aluminum windows, but the price differential steadily increased.
 
Substitution models based on distributed lags of relative p4ces appear to provide more accurate and detailed information on market share changes than models that rely on arbitrary technological innovation formulations. While the short-run own-price and cross-price market share elasticities may be low, the long-run elasticities suggest that direct substitution between competitive products, such as wood and aluminum windows, can exceed 1.0. The case study shows a 1.7 percent change in wood window market share in the long-run for every one percent change in relative price. This high long-run price elasticity of substitution may have significant implications for carbon mitigation analysis and other environmental policy issues.
 
The primary differences between wood and non-wood alternatives used in residential and light commercial construction are the energy requirements involved and carbon emissions related to fossil fuel consumption in production. When all of the aspects of extraction, transportation, processing and production were considered, wood products were found to require less energy in manufacture (CORRIM, 1976). The exact amounts differed for each end-product.
 
Several studies have examined the environmental impacts arising from the production of wood, plastics, aluminum, steel and concrete. Each of these industries has substantial extraction impacts. The manufacturing of steel, aluminum and plastics was judged to create more significant problems than sawn-wood and cement (Alexander and Greber, 1991). Furthermore, wood has the unique attribute of being a renewable resource which can be managed over many rotations.
 
The most significant effect of forest management on the level of carbon dioxide in the atmosphere will come from the substitution of wood materials for more energy-intensive materials. New approaches to forest production and more intensive forest management practices requiring forest investment would be necessary to increase the production of higher-quality logs needed in order to increase the substitution of wood for non-wood products. While shortages of wood products due to forest preservation constraints may reduce wood demand and forest investment on one hand, carbon taxes on fossil fuels could have the impact of increasing demand and forest investment resulting in the substitution of wood for non-wood products on the other. In effect, environmental policies need to consider substitution issues which are likely to show a strong advantage for the use of renewable resources.