1# RECORD 03bfeba917085bc65a0ab641e5a03916b687b20f2# TITLE: Economic feasibility of wood-based structures—Improving urban carbon neutrality strategies3# YEAR: 2021.0 VENUE: nan DOI: 10.1088/2634-4505/abfe054# AUTHORS: I. Talvitie; J. Vimpari; S. Junnila5# PAGES: 10 CANDIDATE TABLE PAGES (1-based): [5, 7]6# ABSTRACT: Countries and cities alike have set carbon neutrality goals to the near future. Urban areas are experiencing an increased demand for new housing developments, and buildings are responsible for one third of all greenhouse gas emissions. The relative share of construction emissions is increasing as energy production is de-carbonizing. Wooden structures have showcased potential in substantially decreasing these emissions. However, for the construction industry to utilize environmentally friendly products, the cost efficiency needs to be determined. Prior research does not present conclusive results regarding the construction costs or value of wooden buildings. This study aims to determine the economic feasibility of wood-based structures in housing. We approach this by estimating the effect of wood on Finnish dwelling prices through hedonic regression analyses. Dwelling prices are analysed with transaction data from the Helsinki metropolitan area. It is provided by the Central Federation of Finnish Real Estate Agencies. The data represents occurred transactions between 1999 and 2018. The study discovers that a wooden structure has a positive price effect in the Helsinki metropolitan region suggesting an economically feasible opportunity to create low carbon urban housing. In the city of Helsinki, the effect is statistically significant (+8.85%) whereas no significance is found in either Espoo or Vantaa. The results suggest that in Helsinki, a direct financial benefit exists for both the city and private developers. For other cities, the study encourages to increase the consumption of wood due to its environmental properties. The study implies that wooden construction is an economically feasible solution in mitigating climate change. As this may provoke an increased demand for wood in construction, further research should be conducted to analyse its effects on the economy and the environment.789===== PAGE 1 (score 1, numeric-paren 0, table-words 0, var-words 1) =====1011 12 13LETTER • OPEN ACCESS14Economic feasibility of wood-based15structures—Improving urban carbon neutrality16strategies17To cite this article: Ilmari Talvitie et al 2021 Environ. Res.: Infrastruct. Sustain. 1 01100218 19View the article online for updates and enhancements.20You may also like21How much upfront-embodied GHG22emissions can wooden buildings23save—displacement factors for wooden24buildings25Ali Amiri and Seppo Junnila26-27Cities as carbon sinks—classification of28wooden buildings29Ali Amiri, Juudit Ottelin, Jaana Sorvari et30al.31-32Benefits of wooden structure reuse: the33case of an Austrian building34E Hoxha, B Soust-Verdaguer, M Scherz et35al.36-37This content was downloaded from IP address 132.203.214.169 on 20/10/2026 at 20:13383940===== PAGE 5 (score 57, numeric-paren 0, table-words 3, var-words 8) =====4142Environ. Res.: Infrastruct. Sustain. 1 (2021) 01100243Table 1. Summary statistics of regression variables.44Summary statistics45HMA all observations46HMA wooden observations47Helsinki all48Helsinki wood49Espoo all50Espoo wood51Vantaa all52Vantaa wood53Mean54Std.dev55Mean56Std.dev57Mean58Std.dev59Mean60Std.dev61Mean62Std.dev63Mean64Std.dev65Mean66Std.dev67Mean68Std.dev69Housing characteristics70Price (€/m2)713,988.40721,419.87733,718.5674932.11754,271.92761,604.94773,730.3778989.22793,800.47801,139.35813,442.9182740.99833,489.8184990.37853,844.4186954.4887Size (m2)8864.168922.269054.899116.749267.359323.939458.009520.139663.339719.509857.949913.7910056.8510119.0810251.5910315.20104Built year1052,005.201067.421072,006.921089.281092,005.001107.341112,006.081127.821132,004.681146.761152,004.391168.521172,006.441188.271192,008.6312010.08121Rooms12211230.091240.281250.141260.351270.071280.261290.181300.381310.071320.251330.081340.281350.151360.361370.151380.3513921400.461410.501420.461430.501440.451450.501460.431470.501480.471490.501500.441510.501520.471530.501540.491550.5015631570.311580.461590.301600.461610.311620.461630.281640.451650.341660.471670.391680.491690.281700.451710.271720.45173Condition174Excellent1750.011760.091770.041780.181790.011800.10181—182—1830.011840.081850.011860.121870.011880.111890.071900.25191Good1920.801930.401940.671950.471960.791970.411980.681990.472000.852010.352020.812030.402040.772050.422060.592070.49208New2090.072100.262110.212120.402130.072140.252150.252160.432170.032180.182190.122200.322210.132220.332230.222240.42225Decent2260.072270.252280.082290.262300.082310.272320.062330.232340.052350.212360.042370.202380.052390.222400.102410.30242Poor2430.002440.03245—246—2470.002480.03249—250—2510.002520.02253—254—2550.002560.04257—258—259Other2600.052610.222620.022630.132640.052650.222660.012670.112680.062690.242700.022710.142720.032730.182740.022750.14276New development (dummy)2770.422780.492790.332800.472810.432820.492830.332840.472850.392860.492870.152880.362890.442900.502910.422920.49293Elevator (dummy)2940.762950.432960.132970.342980.782990.413000.323010.473020.753030.433040.143050.353060.723070.453080.023090.13310Lot ownership311Owned3120.763130.433140.643150.483160.623170.493180.243190.433200.953210.223220.653230.483240.873250.343260.873270.34328Rented3290.233300.423310.353320.483330.373340.483350.733360.453370.033380.183390.343400.473410.133420.343430.133440.33345Other3460.013470.103480.013490.123500.013510.103520.033530.173540.013550.123560.013570.123580.013590.073600.013610.08362Distance to CBD (km)3639.713644.6236513.743664.053676.833683.603699.173702.1337112.003723.0737315.073742.8137514.243763.1437715.783783.19379Sold during or after 20113800.563810.503820.763830.433840.543850.503860.623870.493880.533890.503900.713910.463920.643930.483940.873950.33396Neighbourhood characteristics397Population398517.41399336.68400315.65401222.81402620.57403373.45404507.34405264.40406422.73407262.96408287.83409118.05410370.52411203.51412216.34413152.56414Mean income per capita (€/year)41525,505.544168,625.9041726,929.654187,074.3641926,206.2342010,325.4942123,331.944224,237.3442325,811.184247,087.5442529,949.174267,251.7842723,263.134283,878.3842927,590.664307,450.19431Number of buildings43212.854338.4643423.8443512.5643612.484378.0143820.1743910.1844013.244418.4144226.7744315.1044413.334459.5644624.5944712.02448Mean dwelling size (m2)44964.3745012.4345174.3145216.5345364.3145412.1045567.8645611.7545766.0445814.2445974.1846016.9646162.3446210.2646378.1746417.55465Share of university education4660.254670.144680.384690.154700.254710.144720.364730.124740.254750.154760.464770.184780.254790.114800.344810.13482Share of families with children4830.174840.104850.174860.164870.174880.094890.144900.104910.174920.114930.164940.174950.164960.124970.204980.17499Share of home-owners5000.115010.185020.205030.305040.115050.175060.095070.145080.105090.175100.235110.325120.135130.205140.255150.34516Umemployment rate5170.405180.295190.365200.355210.405220.285230.305240.255250.435260.305270.415280.365290.365300.305310.385320.39533Pensioner share5340.345350.155360.275370.195380.355390.155400.305410.175420.385430.155440.325450.235460.265470.125480.235490.175504551552553===== PAGE 6 (score 23, numeric-paren 0, table-words 9, var-words 5) =====554555Environ. Res.: Infrastruct. Sustain. 1 (2021) 011002556Figure 1. Occurred apartment building dwelling transactions in the HMA.5572.3. Wooden structures in the data558A wooden dwelling in this study is defined as a flat located in a building which is fully constructed of wood or559contains a wooden frame, e.g. buildings with a wooden frame but a brick facade are considered as a wooden560building in this study. Unfortunately, in most cases the data descriptions only detail the main building material561which can lead to incorrect material categorisations for dwellings. The data was processed carefully to avoid562this. The dwelling transaction data is presented in the Finnish language, thus, the following descriptions are563translated into English by the authors. These material descriptions convey the definitions of wooden apartment564building dwellings presented in the data;565(a) ‘Wood’ (obs. count 404)566(b) ‘Wood’ (obs. count 208)567(c) ‘CLT-frame with wooden façade’ (obs. count 11).568Wood’s share of buildings is only 2.23%. The city of Helsinki contains 177 of these observations, the city of569Espoo 145, and the city of Vantaa 301. Figure 1 presents a map that conveys the occurred transactions of apart-570ment building dwellings, which have been constructed between 1990 and 2020 and sold between 1999–2018.571The green diamonds present wood and the white circles non-wooden observations. It is good to note that572the observations are located in multi-storey developments, therefore, each mark in the figure can represent573multiple observations.5743. Results575Table 2 presents regression results based on the model and methods showcased in section 2.1. The table details576the effect of wood, the explanatory power, sample size, and variable-groups used in each of the regression577analyses. The brackets following the variable group ‘postal codes’ indicates the amount of clusters used in the578regression analyses. Each regression uses the same base model, however, the resulting regression models differ579from each other due to the stepwise-method. The statistically significant variables are presented in table 3.580In table 2, the results for the HMA present that a wooden structure has a positive effect (ß-value = 0.017)581on dwelling prices in the entire region. However, based on a 95% confidence interval the effects is not statisti-582cally significant (p = 0.498). The explanatory power, adjusted R-squared, value is nearly 91%. Therefore, the583study does not consider adding additional variables. The regression analysis does not indicate any problems of584multicollinearity (VIF < 10).585To study the HMA in more detail, additional regression analyses were conducted for each city separately.586The reason behind the additional analyses is due to each city having their own urban developmentand housing5875588589590===== PAGE 7 (score 39, numeric-paren 0, table-words 14, var-words 9) =====591592Environ. Res.: Infrastruct. Sustain. 1 (2021) 011002593Table 2. Regression estimates for transaction prices. ∗p-value < 0.05, ∗∗p-value < 0.01, ∗∗∗p-value < 0.001.594(a)–(c) are presented in table 1. [Brackets] convey the amount of postal codes.595(1) HMA596(2) Helsinki597(3) Espoo598(4) Vantaa599Effect of wood6000.0176010.085∗∗∗6020.029603−0.011604Std. err.605[0.025]606[0.020]607[0.037]608[0.022]609R-squared6100.9076110.916120.8916130.876614Adj R-squared6150.9066160.9096170.896180.875619Sample size62027 95462114 86762274226235665624Wooden observations625623626177627145628301629% of wooden obs.6302.23 %6311.19 %6321.95 %6335.31 %634Postal codes635Yes [137]636Yes [79]637Yes [33]638Yes [25]639(a) Housing characteristics640Yes641Yes642Yes643Yes644(b) Neighbourhood characteristics645Yes646Yes647Yes648Yes649(c) Distance to CBD650Yes651Yes652Yes653No654Table 3. Statistically significant variables.655(1) HMA656(2) Helsinki657(3) Espoo658(4) Vantaa659Housing characteristics660Size (log)661Age (log)662Rooms663166426653666Condition667Excellent668Good (control variable)669—670—671—672—673New674Decent675Poor676Other677New development (dummy)678Elevator (dummy)679Lot ownership680Owned681Rented (control variable)682—683—684—685—686Other687Distance to CBD (log)688Sold during or after 2011689Neighbourhood characteristics690Population (log)691Mean income per capita (log)692Number of buildings (log)693Mean dwelling size (log)694Share of university education695Share of families with children696Share of home-owners697Unemployment rate698Pensioner share699programmes. Therefore, individual analyses provide each city with detailed results in regard to future policy700making.701Table 2 presents that a wooden structure has a statistically significant positive effect (p-value = 0.000; ß-702value = 0.085) on dwelling prices in the capital city of Helsinki. The positive effect translates to a 8.87% price703premium for wooden dwellings. For example, considering a non-wooden apartment building which has a704market value of 200 000 €, the equivalent wooden dwelling will have a nearly 18 000 € higher price. Further-705more, the regression analysis’ adjusted R-squared value is nearly 91% and it is clustered by 79 postcodes. The706regression analysis does not indicate any problems of multicollinearity (VIF < 10).7076708709710===== PAGE 8 (score 31, numeric-paren 0, table-words 11, var-words 8) =====711712Environ. Res.: Infrastruct. Sustain. 1 (2021) 011002713Table 4. Alternative regression approach—gradually adding variables.714(1) Helsinki715(2) Helsinki716(3) Helsinki717(4) Helsinki718(5) Helsinki719Wood720−0, 259∗∗721−0, 167∗7220.0437230, 041∗7240, 085∗∗∗725Std. err.726[0, 079]727[0, 07]728[0, 024]729[0, 020]730[0, 020]731R-squared7320.0037330.5747340.857350.8947360.91737Adj R-squared7380.0037390.5747400.857410.8947420.909743N74414 88874514 88874614 88874714 86774814 867749(a) Age & size750Yes751Yes752Yes753Yes754(b) Housing characteristics755Yes756Yes757Yes758(c) CBD Euclidean distance759Yes760Yes761Yes762(d) Neighbourhood characteristics763Yes764Yes765(e) Postal codes766Yes [79]767In Espoo a wooden structure similarly has a positive effect on dwelling prices. However, the effect is not768statistically significant based on a 95% confidence interval (p-value = 0.443). The adjusted R-squared value769is 89% and the model is clustered by 33 postcodes. The regression analysis does not indicate any problems of770multicollinearity (VIF < 10).771The effect of a wooden structure, on the other hand, is negative in the city of Vantaa. However, the effect is772not statistically significant based on a 95% confidence interval (p-value = 0.613). For Vantaa, the explanatory773power of the model is 87.5%, and it is clustered by 25 postcodes. However, the variable ‘distance to centre’ was774excluded from the analysis for Vantaa since it induced high problems of multicollinearity. The study excludes775the variable from the analysis since it decreases the VIF levels considerably and does not alter the regression776results significantly.777In addition to the full regression models, this study presents an alternative approach. Table 4 presents the778alternative manner to analyse the effect of wood on dwelling prices in Helsinki. The figure details an approach779that gradually adds explanatory variables into the regression model. A similar approach is presented in Fuerst780et al (2016). The first model (1) uses the coefficient wood as the only explanatory variable. The second model781adds age and size as additional variables since they have the most effect on the explanatory power of the model.782This is followed by the remaining housing and neighbourhood characteristics. The final model (5) is the same783as presented in table 2. Interestingly, table 4 conveys that when wood is used as the only explanatory variable it784has a statistically significant negative effect. However, by adding more hedonic variables into the equation the785effect of wood at first neutralizes and eventually culminates into a statistically significant positive effect. Each786regression analysis is clustered by 79 postal codes.7874. Discussion and conclusion788The share of global population living in urban areas is increasing rapidly, which stimulates an increased789demand for new urban housing and subsequent GHG emissions. At the same time, cities across the globe have790set carbon neutrality as near future goals. Since carbon neutrality goals are approaching fast, and wood has the791potential to decrease emissions, this study was set to evaluate the economic feasibility of wood construction in792urban settings.793The study found a positive price impact for wooden apartments in the Helsinki metropolitan region. The794positive price premium of 8.87% for wooden dwellings presents a direct financial benefit for the city and private795developers in the capital city of Helsinki. Based on the results, this study suggests the city of Helsinki should796increase the consumption of wood in construction. In addition to the financial benefit, research has showcased797the capabilities of wooden structures in mitigating climate change (Amiri et al 2020 and Buchanan and Levine7981999). These environmental benefits would directly assist in achieving the carbon neutrality goals set by the799city.800For Espoo and Vantaa respectively, the results present that a wooden structure does not affect dwelling801prices significantly in respect to non-wooden materials. Even without a direct financial benefit the study argues802for the two cities to consider wooden construction as a tool to offset emissions. The European Green Deal803(2019) states that to reach a climate neutral future it requires actions such as investing in environmentally-804friendly technologies. Based on prior research (Amiri et al 2020, Buchanan and Levine 1999 and ˇsvajlenka805et al 2017) this study considers wooden construction to be such a technology that can reduce urban emissions.806Therefore, the study proposes that Espoo and Vantaa consider wooden construction materials to be lifted on a807higher pedestal when compared to non-wooden materials, e.g. concrete, due to the material’s environmental808capabilities. Especially with the cities attempting to achieve carbon neutrality by 2030.8097810