## Introduction

The global softwood log market has experienced significant changes over the last decade as a result of fluctuations in market conditions and shifting forest policies. In 2006, Russia exported 37.2 million m3 of softwood logs, which accounted for 45 per cent of world exports (Food and Agriculture Organization of the United Nations (FAO), 2014). However, Russia's softwood log exports began to drop in 2007 following the Russian government's decision to implement an ad valorem log export tax on unprocessed logs as part of an attempt to encourage the growth of its domestic wood processing sector (CIBC World Markets, 2007; Simeone, 2012). In 2012, Russia's softwood log exports declined to ∼12 million m3. Simultaneously, New Zealand benefited from the high prices of Russian logs, and its total softwood log exports increased dramatically – by 83 per cent – from 7.5 to 13.7 million m3 over the period of 2003–2012 (Global Trade Information Services Inc., 2014). New Zealand surpassed Russia as the largest softwood log exporter in the world in 2012 and has the production capacity to expand its timber harvest from plantation forests over the next decade (Ministry of Agriculture and Forestry, 2010).

Despite the significant shifts in Russia's and New Zealand's export trade shares in the global softwood log market, few scholars have analysed the potential effects of the Russian export tax on softwood logs and New Zealand's restrictions on softwood log production on that market. Previous studies have tended to focus solely on examining the quantitative effects of Russian log export restrictions on global log markets. Turner et al. (2008) used a spatial partial equilibrium model to examine the implications of increasing the Russian log export tax for the Russian and global wood product sectors. The authors found that a softwood log export tax of €50 per cubic metre would reduce Russian log exports and prices in 2020 by 50 and 16 per cent, respectively, compared with the base scenario with no tax. In response, New Zealand, the US and Germany would experience increases in both log production and exports. Solberg et al. (2010) employed a spatial partial equilibrium model to analyse the effects of alternative levels of the export tax and variations in the Russian investment climate on the Russian and global wood product sectors through 2020; they found that a unit export tax of €50 per cubic metre would not result in substantially stronger development in the Russian wood processing industry (e.g. the lumber and pulp sectors) than a unit tax of €15 per cubic metre (which was the tax planned by the Russian government). In particular, Russia's forestry and logging industry would be severely affected by a log export tax of €50 per cubic metre, with projected reductions in log production and exports of ∼40 and 100 per cent, respectively, in 2020. Solberg et al. (2010) thus suggested that improving the Russian investment climate would have a greater positive impact on the development of the Russian forest industry than a log export tax.

Although Turner et al. (2008) and Solberg et al. (2010) provide important contributions for policy makers and forest managers who are examining the potential effects of an increased Russian log export tax (e.g. a specific tax of €50 per cubic metre or an 80 per cent ad valorem tax) on the global forest products markets, less is known regarding the potential impacts of recently proposed reductions in the softwood log export tax following Russia's accession to the World Trade Organization (WTO) in 2012. In particular, Russia has established a new trade policy with respect to the softwood log export tax and abandoned its goal of imposing the 80 per cent ad valorem softwood log export tax that was analysed by previous researchers. Instead, Russia established a new volume tariff rate quota system and has agreed to reduce the ad valorem tax rate on softwood logs to the final bound rate of 8 per cent by 2015 for log exports below quota. For exports above the quota, a 25 per cent export tax might be applied (Random Lengths International, 2012; Simeone and Eastin, 2012).

Softwood logs are New Zealand's largest forest products export in terms of both value and volume (Ministry for Primary Industries, 2014). New Zealand's supply of softwood logs is mainly drawn from 1.8 million ha of highly productive plantation forests. Statistical data have revealed that New Zealand's annual log production in 2012 was 26.1 million m3 and that ∼50 per cent of that production was exported in log form (Ministry for Primary Industries, 2014). It is projected that the softwood log production from plantation forests might be expanded up to 30 or even 35 million m3 per year in the 2020s (Ministry of Agriculture and Forestry, 2010). Because plantation forest estates account for >99 per cent of New Zealand's total timber harvest and >90 per cent of these estates are privately owned, investment in new plantations and harvesting decisions generally depend on market and economic conditions, including log prices, shipping costs and currency fluctuations. However, there are certain social and environmental factors that might also potentially impact future timber harvest decisions and softwood log supplies in New Zealand, including the increase in aboriginal (Māori) ownership of forests and the implementation of emissions trading schemes (Approximately 13 per cent (238 000 ha) of New Zealand's plantation forests are owned by the Māori people, and this ownership share of plantation forest areas is expected to increase to as high as 41 per cent in the future in accordance with anticipated treaty claim settlements (Miller et al. 2007)) (Rhodes and Novis, 2004; Miller et al., 2007; Ministry of Agriculture and Forestry, 2010).

The purpose of this study is to utilize a trade flow model to forecast the trade flow trend of global softwood logs and also to examine the potential economic impact on the world market of proposed forest policies in both Russia and New Zealand. A recursive dynamic world spatial partial equilibrium model was first used to project baseline production, consumption, trade flow volumes and prices of softwood logs through 2021 and then compare them with the results of the following three alternative scenarios: (1) Russia reduces its ad valorem softwood log export tax to 8 per cent and also removes the volume tariff rate quota system to comply with the final bound rates established in relevant WTO accession agreements; (2) New Zealand does not expand timber harvest in plantation forests due to social and environmental considerations (e.g. increasing Māori ownership of forest land and the implementation of emissions trading schemes), which leads to zero growth in the New Zealand softwood log production from 2011 to 2021; and (3) a combination of the proposed policies in Russia and New Zealand mentioned above. The results of this study will provide insights for forest managers and policy makers to examine the potential effects of changes in forest policies in these two important softwood log supply regions on the global softwood log market.

## Research method and materials

The spatial partial equilibrium model applied in this study was originally formulated by Samuelson (1952) and further developed by Takayama and Judge (1971) and has been used frequently by researchers to investigate the effects of trade policy changes and/or market shocks on the quantities and prices of forest products (Latta et al., 2013). The model used in this study is similar to the softwood lumber trade flow models developed by Gaston et al. (1999); Gaston and Marinescu (2006) and Chang and Gaston (2014). The model considers several net demand and net supply regions that are separated spatially and also considers transportation costs in determining competitive prices and quantities that will maximize the total trade surplus of the market under several linear material balance constraints. In this study, each region is considered to be either a net demand or net supply region for softwood logs with the following inverse net (excess) demand or net supply function in the objective function, which were derived from the domestic demand and supply functions in each region (See Appendix A for the derivation of the net demand (or net supply) function from the domestic demand and supply functions in each region.):

(1)
$Pid=αi−βiMi,i=1,…,n$

(2)
$PjS=γj+δjXj,j=1,…,m$
where αi and βi denote the intercept and slope, respectively, of the net demand function of softwood logs and the variables $Pid$ and Mi represent the demand (import) price and total quantity demanded (imports), respectively, for net demand region i. For net supply region j, γj and δj are the intercept and slope of the net supply function of softwood logs, respectively, and $PjS$ and Xj are the supply (export) price and total quantity supplied (exports), respectively. The objective of the model is to maximize the trade surplus in all regions, which is the sum of all regional demand integrals less the sum of all regional supply integrals and interregional transportation costs. Because the net demand and supply functions are assumed to be linear functional forms, the integrals of the objective function can be expressed in terms of the quadratic function in equation (3) subject to the related material balance constraints in equations (4–7) below:
(3)
$Max∑i=1nαiMi−12βi(Mi)2−∑j=1mγjXj+12δj(Xj)2−∑i=1n∑j=1mtijQij$

(4)
$∑i=1nQij≤Xj∀j$

(5)
$∑j=1mQij≥Mi∀i$

(6)
$Mi,Xj,Q ij≥0∀i,j$

(7)
$∑i=1nMi−∑j=1mXj=0$
where the first and second terms of equation (3) represent the sum of importer and exporter trade surpluses for softwood logs, respectively. The third term is the sum of interregional transportation costs. Qij is the quantity of softwood logs exported from region j to region i, and tij is the per-unit transport cost of softwood logs from region j to region i. Equation (4) ensures that the total exports from net supply region j are equal to or less than the quantity supplied in region j. Equation (5) ensures that the total quantity imported into net demand region i is greater than or equal to the quantity demanded in region i. Equation 6 states that the quantities of demand, supply and shipments are non-negative. The last constraint (equation 7) presents the world market equilibrium conditions.

### Countries/regions

The spatial partial equilibrium model developed for this study included 13 countries/regions, of which seven are net supply regions and six are net demand regions. The net supply regions consist of Canada, the US export region (mainly the western US), the European export region (including Estonia, Latvia, Lithuania, Germany, Austria, Finland and Sweden), the Russian Federation, New Zealand, Chile and the rest of the world export region. The net demand regions consist of the US import region (mainly the southern and northern US), Japan, Korea, China, the European import region (including Norway, Switzerland and the European Union-27 countries but excluding the European export countries listed above) and the rest of the world import region.

### Data used in the spatial partial equilibrium model

Data employed in the spatial partial equilibrium model require base year data on softwood log trade flows, production, consumption, prices, own-price elasticities of supply and demand and transportation costs for each region in order to derive the required parameters (e.g. intercepts and slopes) of domestic demand and supply functions and also the net demand (or net supply) functions (see Appendix A). The base year (2011) trade flow data that we used in the spatial partial equilibrium model were sourced from the Global Trade Atlas® (Global Trade Information Services Inc., 2014). Data on softwood log production were obtained from Natural Resources Canada (2012), FAO (2014), and Oswalt et al. (2014). Each net supply (export) region's consumption of softwood logs equalled production minus exports, whereas each net demand (import) region's consumption equalled production plus imports. In addition, the base year prices of softwood logs were derived from the weighted averages of the unit values of exports or imports appearing in the trade flow data of the Global Trade Atlas®. The domestic price elasticities of supply and demand for softwood logs range from 0.8 to 1.6 and from −0.3 to −0.4, respectively, which are adopted from Abbott et al. (2009) and Buongiorno and Zhu (2013).

Table 1 lists the transportation costs used in the model and were based on lumber transportation costs and inflated by 4/π to account for the space difference between logs and lumber, following Abbott et al. (2009). For the sake of simplicity and also to more easily distinguish among effects of different policy scenarios, we assumed that the transportation costs (in real dollar value terms) for softwood logs would remain the same over the projection period (Depending on the research needs and interests, the model can easily substitute basic data and run the program in Excel to determine the sensitivity to assumptions about price elasticities, transportation costs, etc.). In addition, because transportation costs in this study were not calibrated to the price differences between two regions, following Paris et al. (2011) and van Kooten and Johnston (2014), positive mathematical programming method was used to generate the dual values (shadow prices) that can be used to adjust original transportation costs (tij in the objective function) to represent the real transaction costs between export and import regions. Using this method allows those factors that are not included in transportation costs (e.g. heterogeneous log quality, tariff/non-tariff barriers, export subsidies, decision makers' risk, etc.) to be considered in the model (From a different perspective, since the model does not solely rely on original transportation costs (estimated by the shipping costs between regions) in the solution process but uses real transaction costs estimated by positive mathematical programming method to determine the optimal quantities and prices among regions, this may also mitigate the sensitivity of original transportation costs in traditional spatial partial equilibrium modelling. Further research is warranted).

Table 1

Transportation costs for softwood logs used in the spatial partial equilibrium model (US$/m3) Export region Import region US Japan Korea China Europea Rest of world Canada 48 64 64 64 79 101 US 64 64 64 58 99 Europeb 30 67 67 67 17 64 Russia 55 28 28 28 17 64 New Zealand 87 54 54 54 113 73 Chile 51 106 106 106 72 49 Rest of World 77 91 80 80 60 Export region Import region US Japan Korea China Europea Rest of world Canada 48 64 64 64 79 101 US 64 64 64 58 99 Europeb 30 67 67 67 17 64 Russia 55 28 28 28 17 64 New Zealand 87 54 54 54 113 73 Chile 51 106 106 106 72 49 Rest of World 77 91 80 80 60 The transportation costs were based on the lumber transportation costs from Chang and Gaston (2014) and inflated by 4/π to account for the space difference between logs and lumber, following Abbott et al. (2009). The US import region was mainly the southern and northern US, whereas the US export region was mainly the western US. aIncluding Norway, Switzerland and the European Union (27 countries other than the European export countries included in footnote b). bIncluding Germany, Austria, Finland, Sweden, Estonia, Latvia and Lithuania. ### Baseline projections of future supply and demand of softwood logs To project the future supply and demand of softwood logs over time (2012–2021), a recursive dynamic procedure was employed in the model to forecast baseline production, consumption, trade volumes and prices for each region at different times. In particular, the intercepts of the domestic supply and demand functions for each region were changed exogenously according to the expected (or assumed) annual changes in log supply and demand, and the input parameters were then updated recursively for each 5-year period. Therefore, the nature of the model assumes that the decision makers in the economy have imperfect foresight (Solberg et al. (2010)). Factors that may influence future softwood log supply and demand, including economic growth, population, technology, investment, harvest inventory and government policy, could therefore be included in the model as intercept shifters (see Appendix B). The expected annual changes in softwood log supply and demand for each region over the 2012–2021 period are shown in Table 2 and were taken from Sedjo (1999); British Columbia Ministry of Forests and Range (2007); FAO (2009); British Columbia Ministry of Forests, Lands, and Natural Resource Operations (2012); Oswalt et al. (2014), and the authors' judgement. Some of the anticipated changes (or assumptions) that will affect global softwood log supply and demand include the following: (1) Annual demand in the US would increase by 2 per cent due to the recovery of the American housing market; (2) annual supply in the US import region would increase by 2 per cent due to softwood plantations in the southern US; (3) production in New Zealand and Chile would grow by 2 per cent annually due to well-stocked plantation forests; (4) annual timber harvests in Russia would be affected by the current softwood log export tax, and we assumed that the annual softwood log production in Russia would thus decrease by 1 and 0.5 per cent between 2012 and 2016 and between 2017 and 2021, respectively and (5) the timber harvest in Canada would be affected by the current mountain pine beetle infestation in the province of British Columbia, and we estimated that the annual production of softwood logs in Canada would decrease by 15 million cubic metres over the 2012–2021 period, which is equal to annual reduction rates of 2.1 and 3.5 per cent between 2012 and 2016 and between 2017 and 2021, respectively. Table 2 Assumptions about future average annual supply and demand changes (%) for softwood logs Region 2012–2016 2017–2021 Supply Demand Supply Demand Export Canada −2.1 0.0 −3.5 0.0 US 0.0 2.0 0.0 2.0 Europea 1.0 1.5 1.0 1.5 Russia −1.0 1.0 −0.5 1.0 NZ 2.0 0.0 2.0 0.0 Chile 2.0 1.0 2.0 1.0 Rest of world 1.0 1.7 1.0 1.7 Import US 2.0 2.0 2.0 2.0 Japan 1.3 0.0 1.3 0.0 Korea 1.3 0.0 1.3 0.0 China 1.3 2.0 1.3 2.0 Europeb 1.0 1.5 1.0 1.5 Rest of world 1.8 1.7 1.8 1.7 Region 2012–2016 2017–2021 Supply Demand Supply Demand Export Canada −2.1 0.0 −3.5 0.0 US 0.0 2.0 0.0 2.0 Europea 1.0 1.5 1.0 1.5 Russia −1.0 1.0 −0.5 1.0 NZ 2.0 0.0 2.0 0.0 Chile 2.0 1.0 2.0 1.0 Rest of world 1.0 1.7 1.0 1.7 Import US 2.0 2.0 2.0 2.0 Japan 1.3 0.0 1.3 0.0 Korea 1.3 0.0 1.3 0.0 China 1.3 2.0 1.3 2.0 Europeb 1.0 1.5 1.0 1.5 Rest of world 1.8 1.7 1.8 1.7 The assumptions are based on reported data from Sedjo (1999); British Columbia Ministry of Forests and Range (2007); FAO (2009); British Columbia Ministry of Forests, Lands, and Natural Resource Operations (2012);Oswalt et al. (2014), and the authors' judgement. The US export region was mainly the western US, whereas the US import region was mainly the southern and northern US. aIncluding Germany, Austria, Finland, Sweden, Estonia, Latvia and Lithuania. bIncluding Norway, Switzerland and the European Union (27 countries other than the European export countries included in footnote a). ### Scenario analysis The baseline projection was compared with three independent simulations involving proposed forest policy scenarios in Russia and New Zealand, including the following: (1) Russia reduces its ad valorem softwood log export tax to 8 per cent and also removes the volume tariff rate quota system to comply with the final bound rates established in the relevant WTO accession agreements (See Appendix C for the mathematical derivation of changes in net supply function when an ad valorem export tax is considered in the trade flow model); (2) New Zealand does not expand its timber harvest in plantation forests because of social and environmental considerations, with no assumed growth (0 per cent) in the annual softwood log production over the 2011–2021 period; and (3) a combination of the proposed policies in Russia and New Zealand above. The equations needed in the spatial partial equilibrium model, such as the domestic demand and supply functions, derived inverse excess demand and supply functions, the objective function and constraints were written in Microsoft Excel 2013 and solved using the third-party add-in solver What's Best version 12.0 (Lindo System Inc., 2013). The What's Best solver retrieved the data directly from Excel, calculated the optimized solution for each period, and then returned the results to Excel. The model was first solved by using a positive mathematical programming method to calibrate to observed softwood log production, consumption, trade volumes and prices in the base year 2011 and then made dynamic by shifting the intercepts of domestic demand and supply functions inter-periodically to reflect the exogenous assumptions of future supply and demand of softwood log conditions in each region through 2021 (see Table 2 and Appendix B). For each scenario, optimal trade quantities and prices were estimated recursively every 5 years over the forecast period to allow enough time for the global softwood log market to adjust to the proposed policy changes, and the equilibrium for each period was updated according to the supply and demand conditions from the previous period (The model can also project the optimal trade quantities and prices every year by a recursive dynamic procedure. However, this may require more detailed assumptions of expected annual supply and demand changes in each year). The major economic variables of interest in the analysis included the effects on future softwood log prices, production, consumption and trade flow volumes for each region. ## Results ### Baseline projections Table 3 presents calibrated softwood log production, consumption, trade and prices for the trade flow model in the base year of 2011. The production, consumption, trade and price levels of 2011 are precisely duplicated using positive mathematical programming (Because the intercepts and slopes of the inverse net demand or net supply functions in the objective function were derived from the parameters of the domestic demand and supply functions in each region, where the parameters of domestic demand and supply functions in each region were derived from base year data of production, consumption, trade and prices, the optimal solution of the model by using positive mathematical programming method should duplicate (or forecast) the exact values of observed base year data). Thus, these figures provide a good foundation to be used in the model for projecting global softwood log market conditions during the 2012–2021 period. Table 3 Calibrated softwood log production, consumption, trade and prices for the trade flow model and comparisons with actual levels in base year 2011 Region Production Consumption Trade Price Million m3 % of Actual Million m3 % of Actual Million m3 % of Actual US$/m3 % of Actual
Export
Canada 117.29 100.0 111.87 100.0 5.42 100.0 99 100.0
US 63.85 100.0 51.33 100.0 12.52 100.0 118 100.0
Europe 164.92 100.0 161.70 100.0 3.22 100.0 103 100.0
Russia 100.65 100.0 83.41 100.0 17.25 100.0 97 100.0
NZ 24.21 100.0 11.77 100.0 12.44 100.0 110 100.0
Chile 26.21 100.0 25.17 100.0 1.04 100.0 70 100.0
ROW 62.89 100.0 59.56 100.0 3.33 100.0 94 100.0
Import
US 161.85 100.0 162.47 100.0 0.62 100.0 133 100.0
Japan 15.93 100.0 20.08 100.0 4.15 100.0 213 100.0
Korea 2.37 100.0 8.21 100.0 5.84 100.0 189 100.0
China 64.10 100.0 98.03 100.0 33.93 100.0 151 100.0
Europe 109.73 100.0 113.30 100.0 3.57 100.0 120 100.0
ROW 44.85 100.0 51.95 100.0 7.10 100.0 136 100.0
Region Production

Consumption

Price

Million m3 % of Actual Million m3 % of Actual Million m3 % of Actual US/m3 % of Actual Export Canada 117.29 100.0 111.87 100.0 5.42 100.0 99 100.0 US 63.85 100.0 51.33 100.0 12.52 100.0 118 100.0 Europe 164.92 100.0 161.70 100.0 3.22 100.0 103 100.0 Russia 100.65 100.0 83.41 100.0 17.25 100.0 97 100.0 NZ 24.21 100.0 11.77 100.0 12.44 100.0 110 100.0 Chile 26.21 100.0 25.17 100.0 1.04 100.0 70 100.0 ROW 62.89 100.0 59.56 100.0 3.33 100.0 94 100.0 Import US 161.85 100.0 162.47 100.0 0.62 100.0 133 100.0 Japan 15.93 100.0 20.08 100.0 4.15 100.0 213 100.0 Korea 2.37 100.0 8.21 100.0 5.84 100.0 189 100.0 China 64.10 100.0 98.03 100.0 33.93 100.0 151 100.0 Europe 109.73 100.0 113.30 100.0 3.57 100.0 120 100.0 ROW 44.85 100.0 51.95 100.0 7.10 100.0 136 100.0 Table 4 presents the projected baseline annual production, consumption, trade and prices of softwood logs for 2016 and 2021. These results were based on the expected changes in softwood log supply and demand and aligned with anticipated outcomes. For example, the annual production of softwood logs in Canada was projected to decrease significantly between 2011 and 2021 due to the impact of the mountain pine beetle infestation with production decreasing by 11 per cent (13.4 million m3) over the 10-year period. In addition, the production of softwood logs was projected to significantly increase by 25 per cent (6 million m3) in New Zealand, by 31 per cent (8.1 million m3) in Chile and by 21 per cent (34.6 million m3) in the US import region as a result of increases in plantation forests in these regions. Table 4 Baseline projection of annual production, consumption, net exports/imports and prices of softwood logs for 2016 and 2021 and changes over time (2011–2021) Regions Production (Million M3) Consumption (Million M3) Net Exports or Imports (Million M3) Prices (US/m3)

2016 2021 2011–2021a 2016 2021 2011–2021a 2016 2021 2011–2021a 2016 2021 2011–2021a
Export
Canada 109.76 103.88 −13.41 109.76 103.88 −7.99 0.00 0.00 −5.42 104.5 118.7 19.2
US 66.17 67.96 4.11 56.03 61.30 9.97 10.14 6.66 −5.86 121.2 123.2 4.8
Europe 178.76 192.00 27.08 172.26 183.98 22.28 6.50 8.02 4.80 105.9 108.0 4.8
Russia 99.80 100.52 −0.14 86.73 90.41 7.00 13.07 10.11 −7.14 100.1 102.2 4.8
New Zealand 27.11 30.18 5.97 11.66 11.58 −0.19 15.45 18.61 6.16 112.6 114.6 4.8
Chile 30.24 34.35 8.14 26.07 27.10 1.93 4.17 7.25 6.21 72.8 74.8 4.8
Rest of world 68.57 73.95 11.05 63.99 68.95 9.38 4.58 5.00 1.67 96.4 98.5 4.8
Import
US 178.59 196.45 34.60 178.59 196.45 33.98 0.00 0.00 −0.62 133.1 133.1 0.3
Japan 17.13 18.36 2.43 19.99 19.92 −0.17 2.86 1.55 −2.60 215.6 217.6 4.8
Korea 2.57 2.77 0.40 8.16 8.13 −0.08 5.60 5.36 −0.48 191.7 193.8 4.8
China 69.87 75.64 11.54 107.19 117.41 19.38 37.32 41.77 7.83 153.6 155.6 4.8
Europe 118.57 127.08 17.36 120.85 129.19 15.89 2.28 2.11 −1.46 122.4 124.5 4.8
Rest of world 50.13 55.60 10.75 55.99 60.45 8.50 5.86 4.85 −2.25 138.7 140.8 4.8
Regions Production (Million M3)

Consumption (Million M3)

Net Exports or Imports (Million M3)

Prices (US$/m3) 2016 2021 2011–2021a 2016 2021 2011–2021a 2016 2021 2011–2021a 2016 2021 2011–2021a Export Canada 109.76 103.88 −13.41 109.76 103.88 −7.99 0.00 0.00 −5.42 104.5 118.7 19.2 US 66.17 67.96 4.11 56.03 61.30 9.97 10.14 6.66 −5.86 121.2 123.2 4.8 Europe 178.76 192.00 27.08 172.26 183.98 22.28 6.50 8.02 4.80 105.9 108.0 4.8 Russia 99.80 100.52 −0.14 86.73 90.41 7.00 13.07 10.11 −7.14 100.1 102.2 4.8 New Zealand 27.11 30.18 5.97 11.66 11.58 −0.19 15.45 18.61 6.16 112.6 114.6 4.8 Chile 30.24 34.35 8.14 26.07 27.10 1.93 4.17 7.25 6.21 72.8 74.8 4.8 Rest of world 68.57 73.95 11.05 63.99 68.95 9.38 4.58 5.00 1.67 96.4 98.5 4.8 Import US 178.59 196.45 34.60 178.59 196.45 33.98 0.00 0.00 −0.62 133.1 133.1 0.3 Japan 17.13 18.36 2.43 19.99 19.92 −0.17 2.86 1.55 −2.60 215.6 217.6 4.8 Korea 2.57 2.77 0.40 8.16 8.13 −0.08 5.60 5.36 −0.48 191.7 193.8 4.8 China 69.87 75.64 11.54 107.19 117.41 19.38 37.32 41.77 7.83 153.6 155.6 4.8 Europe 118.57 127.08 17.36 120.85 129.19 15.89 2.28 2.11 −1.46 122.4 124.5 4.8 Rest of world 50.13 55.60 10.75 55.99 60.45 8.50 5.86 4.85 −2.25 138.7 140.8 4.8 aRelated data in 2011 can be found in Table 3. Annual consumption of softwood logs from 2011 to 2021 was projected to increase by 19 per cent (10 million m3) and 21 per cent (34 million m3) in the US export and import regions, respectively, by 14 per cent (22.3 and 15.9 million m3) in the European export and import regions, respectively, by 8 per cent (7 million m3) in Russia, and by 20 per cent (19.4 million m3) in China during the 2011–2021 period due to growth in domestic demand. However, consumption was also projected to decrease by 7 per cent (8 million m3) in Canada and marginally decrease (by less than 1 per cent) in New Zealand, Japan and Korea over the same period. Regarding forecasted total trade (net exports and imports), the model projected that annual total exports of softwood logs in Canada would drop to zero (from 5.4 million m3) as a result of reduced timber supply resulting from the mountain pine beetle infestation and the assumption of increased demand for wood products in the US. The annual total trade in softwood logs in the US would decline by 47 per cent (5.9 million m3) and 100 per cent (0.6 million m3) in the US export and import regions, respectively, based on the growth of domestic timber demand. Softwood log exports in Russia would decrease by 41 per cent (7.1 million m3) because of growth in domestic timber consumption. In addition, annual total exports in Chile and the European export region were projected to significantly increase by 597 per cent (6.2 million m3) and 149 per cent (4.8 million m3) from 2011 to 2021, respectively. Annual total imports in China were projected to increase by 23 per cent (7.8 million m3) from 2011 to 2021. The baseline projection for softwood log prices over the 2011–2021 period revealed that prices would increase by$0.3 to 19.2/m3 in every region. For instance, the prices were projected to increase by $19.2/m3 in Canada and by$4.8/m3 in Russia, New Zealand, Japan and China and to marginally increase by $0.3/m3 in the US import region. The forecasted baseline annual trade flows of softwood logs over the 2011–2021 period are presented in Figure 1. China would continue to be the largest importer of softwood logs in the world; however, the major sources of supply for imports would change significantly over the 2011–2021 period. The model forecasts that Canada will not have extra logs available for export to China and other import regions because of the impact of the mountain pine beetle infestation in western Canada. Annual exports from the US export region to China would also decrease significantly (from 5.4 to 2.8 million m3) over the same period due to increased domestic timber demand. In addition, a significant decrease in exports was forecast from Russia to China (from 14.2 to 8.1 million m3 per year) as a result of the existing softwood log export taxes. To meet the growing domestic timber demand and to compensate for the significant decline in imports from Russia, Canada and the US export region, China would source its imports from the European export region, New Zealand, Chile and the rest of the world export region, with annual imports from those regions projected to significantly increase, i.e. from 0.3 to 5.6 million m3, from 9.6 million to 16.6 million m3, from 0.3 to 5.4 million m3, and from 1.5 to 3.3 million m3 over the 2011–2021 period, respectively. Figure 1 Forecasted baseline trade flows (million m3) for (a) Canada Exports, (b) US Exports, (c) Europe Exports, (d) Russian Federation exports, (e) New Zealand exports, (f) Chile exports and (g) Other Exports (rest of world (ROW)) of softwood logs for 2011, 2016 and 2021. The import regions are on the x-axis. Figure 1 Forecasted baseline trade flows (million m3) for (a) Canada Exports, (b) US Exports, (c) Europe Exports, (d) Russian Federation exports, (e) New Zealand exports, (f) Chile exports and (g) Other Exports (rest of world (ROW)) of softwood logs for 2011, 2016 and 2021. The import regions are on the x-axis. ### Scenario results #### Reduction in Russian softwood log export taxes The effects of reducing the Russian ad valorem softwood log export tax on annual softwood log production, consumption, trade and prices in 2016 and 2021 are presented in Table 5. The softwood log price would increase by$14.0/m3 in Russia over the projection period. Because log prices would rise in Russia, its log production would increase significantly – by 21.6 million m3 – and its consumption would decline by 4.4 million m3, leading to significant export increases of 26.0 million m3 in 2021. However, trade liberalization in Russia would also lead to price declines in other regions of the world, which would drive export regions to reduce their exports and import regions to increase their imports. Moreover, when Russia reduces its softwood log export taxes (Figure 2), considerable reallocation of trade flows in the global softwood log market results. Specifically, Russia would significantly increase its log exports to China and the European import region, with annual softwood log exports projected to increase by 21.4 and 1.8 million m3, respectively, in 2021. Due to the significant increase in imports from Russia, China would reduce its annual imports from the US export region by 2.8 million m3, the European export region by 5.6 million m3, New Zealand by 4.4 million m3, Chile by 3.4 million m3 and the other export regions by 2.9 million m3 in 2021. Overall, the total log exports from Russia would significantly increase by 26.0 million m3, and total trade volume (both exports and imports) in the world would increase by 9.3 million m3 in 2021 in response to the reduction in Russian softwood log export taxes.

Table 5

Impact on annual production, consumption, trade (net exports or imports) and prices as a result of reducing the Russian softwood log export tax in 2016 and 2021a

Production (million m3Consumption (million m3Trade (million m3Price (US$) 2016 Export Canada 0.00 0.00 0.00 0.00 US −2.56 0.48 −3.03 −2.07 Europe −4.30 1.20 −5.50 −2.96 Russia 21.00 −4.29 25.30 13.55 New Zealand −0.52 0.12 −0.64 −2.96 Chile −1.55 0.39 −1.95 −2.96 Rest of World −2.78 0.70 −3.48 −2.96 Import US −1.08 0.25 1.32 −0.55 Japan −0.18 0.10 0.28 −2.96 Korea −0.05 0.05 0.10 −2.96 China −1.76 0.71 2.48 −2.96 Europe −3.69 1.04 4.73 −2.96 Rest of World −1.37 0.42 1.79 −2.96 Total 1.16 1.16 10.70 n/a 2021 Export Canada 0.00 0.00 0.00 0.00 US −2.49 0.46 −2.96 −2.89 Europe −6.09 1.70 −7.79 −2.93 Russia 21.61 −4.42 26.02 13.94 New Zealand −0.52 0.12 −0.63 −2.93 Chile −1.54 0.39 −1.93 −2.93 Rest of world −2.75 0.69 −3.44 −2.93 Import US 0.00 0.00 0.00 0.00 Japan −0.18 0.10 0.28 −2.93 Korea −0.05 0.05 0.10 −2.93 China −1.74 0.70 2.45 −2.93 Europe −3.65 1.03 4.68 −2.93 Rest of world −1.35 0.41 1.77 −2.93 Total 1.24 1.24 9.27 n/a Production (million m3Consumption (million m3Trade (million m3Price (US$)
2016
Export
US −2.56 0.48 −3.03 −2.07
Europe −4.30 1.20 −5.50 −2.96
Russia 21.00 −4.29 25.30 13.55
New Zealand −0.52 0.12 −0.64 −2.96
Chile −1.55 0.39 −1.95 −2.96
Rest of World −2.78 0.70 −3.48 −2.96
Import
US −1.08 0.25 1.32 −0.55
Japan −0.18 0.10 0.28 −2.96
Korea −0.05 0.05 0.10 −2.96
China −1.76 0.71 2.48 −2.96
Europe −3.69 1.04 4.73 −2.96
Rest of World −1.37 0.42 1.79 −2.96
Total 1.16 1.16 10.70 n/a
2021
Export
US −2.49 0.46 −2.96 −2.89
Europe −6.09 1.70 −7.79 −2.93
Russia 21.61 −4.42 26.02 13.94
New Zealand −0.52 0.12 −0.63 −2.93
Chile −1.54 0.39 −1.93 −2.93
Rest of world −2.75 0.69 −3.44 −2.93
Import
US 0.00 0.00 0.00 0.00
Japan −0.18 0.10 0.28 −2.93
Korea −0.05 0.05 0.10 −2.93
China −1.74 0.70 2.45 −2.93
Europe −3.65 1.03 4.68 −2.93
Rest of world −1.35 0.41 1.77 −2.93
Total 1.24 1.24 9.27 n/a

aThis scenario simulated Russia reducing its softwood log export tax to 8% in 2015 and also removing the volume tariff rate quota system to comply with the final bound rates as established in its WTO accession package.

Figure 2

Impact on annual softwood log trade flows (million m3) as a result of reducing the Russian log export tax following World Trade Organization accession in 2016 and 2021 for (a) Canada exports, (b) US Exports, (c) Europe Exports, (d) Russian Federation Exports, (e) New Zealand Exports, (f) Chile exports and (Other Exports (rest of world (ROW)). The import regions are on the x-axis.

Figure 2

Impact on annual softwood log trade flows (million m3) as a result of reducing the Russian log export tax following World Trade Organization accession in 2016 and 2021 for (a) Canada exports, (b) US Exports, (c) Europe Exports, (d) Russian Federation Exports, (e) New Zealand Exports, (f) Chile exports and (Other Exports (rest of world (ROW)). The import regions are on the x-axis.

#### Restriction of New Zealand softwood log production

Table 6 presents the effects of restricting New Zealand softwood log production on annual production, consumption, trade and prices in 2016 and 2021. The model projected that softwood log prices would marginally increase by $0.5/m3 in every region in 2021 (except that there would be no changes in Canada and the US import region). Annual production in New Zealand would drop by 5.1 million m3 in 2021 under this scenario, and its log exports would decrease to a similar extent in 2021 compared with the baseline projection (Table 6). The projected trade flow analysis revealed that annual exports to China from New Zealand would decrease by 4.8 million m3 in 2021 (Figure 3). In response to reduced imports from New Zealand, China would look to other regions for its log imports. Thus, it is projected that China would increase its log imports in 2021 by 0.6 million m3 from the US export region, by 1.5 million m3 from the European export region, by 1.1 million m3 from Russia, by 0.5 million m3 from Chile and by 0.7 million m3 from other export regions in the rest of the world to meet its log demand. Overall, the annual total log exports from New Zealand would decrease by 5.0 million m3 and the annual total trade volume in the world would drop by 1.5 million m3 in 2021 in response to restrictions on softwood log production in New Zealand. Table 6 Impact on annual production, consumption, trade (net exports or imports) and prices as a result of restricting New Zealand's softwood log production in 2016 and 2021a Production (million m3Consumption (million m3Trade (million m3Price (US$)
2016
Export
US 0.19 −0.04 0.23 0.23
Europe 0.47 −0.13 0.60 0.23
Russia 0.35 −0.07 0.42 0.23
New Zealand −2.38 −0.01 −2.37 0.23
Chile 0.12 −0.03 0.15 0.23
Rest of world 0.21 −0.05 0.26 0.23
Import
US 0.00 0.00 0.00 0.00
Japan 0.01 −0.01 −0.02 0.23
Korea 0.00 0.00 −0.01 0.23
China 0.13 −0.05 −0.19 0.23
Europe 0.28 −0.08 −0.36 0.23
Rest of world 0.10 −0.03 −0.14 0.23
Total −0.51 −0.51 −0.71 n/a
2021
Export
US 0.40 −0.08 0.48 0.47
Europe 0.99 −0.28 1.27 0.47
Russia 0.72 −0.15 0.87 0.47
New Zealand −5.05 −0.02 −5.03 0.47
Chile 0.26 −0.06 0.32 0.47
Rest of world 0.45 −0.11 0.56 0.47
Import
US 0.00 0.00 0.00 0.00
Japan 0.03 −0.02 −0.05 0.47
Korea 0.01 −0.01 −0.02 0.47
China 0.29 −0.12 −0.40 0.47
Europe 0.60 −0.17 −0.77 0.47
Rest of world 0.22 −0.07 −0.29 0.47
Total −1.09 −1.09 −1.52 n/a
Production (million m3Consumption (million m3Trade (million m3Price (US$) 2016 Export Canada 0.00 0.00 0.00 0.00 US 0.19 −0.04 0.23 0.23 Europe 0.47 −0.13 0.60 0.23 Russia 0.35 −0.07 0.42 0.23 New Zealand −2.38 −0.01 −2.37 0.23 Chile 0.12 −0.03 0.15 0.23 Rest of world 0.21 −0.05 0.26 0.23 Import US 0.00 0.00 0.00 0.00 Japan 0.01 −0.01 −0.02 0.23 Korea 0.00 0.00 −0.01 0.23 China 0.13 −0.05 −0.19 0.23 Europe 0.28 −0.08 −0.36 0.23 Rest of world 0.10 −0.03 −0.14 0.23 Total −0.51 −0.51 −0.71 n/a 2021 Export Canada 0.00 0.00 0.00 0.00 US 0.40 −0.08 0.48 0.47 Europe 0.99 −0.28 1.27 0.47 Russia 0.72 −0.15 0.87 0.47 New Zealand −5.05 −0.02 −5.03 0.47 Chile 0.26 −0.06 0.32 0.47 Rest of world 0.45 −0.11 0.56 0.47 Import US 0.00 0.00 0.00 0.00 Japan 0.03 −0.02 −0.05 0.47 Korea 0.01 −0.01 −0.02 0.47 China 0.29 −0.12 −0.40 0.47 Europe 0.60 −0.17 −0.77 0.47 Rest of world 0.22 −0.07 −0.29 0.47 Total −1.09 −1.09 −1.52 n/a aThis scenario simulated New Zealand not expanding softwood log production in plantation forests and thus projected a 0% annual timber supply growth rate from 2011 to 2021 due to social and environmental considerations. Figure 3 Impact on annual softwood log trade flows (million m3) as a result of restricting New Zealand's softwood log production in 2016 and 2021 for (a) Canada Exports, (b) US Exports, (c) Europe Exports, (d) Russian Federation Exports, (e) New Zealand Exports, (f) Chile Exports and (g) Other Exports (rest of world (ROW)). The import regions are on the x-axis. Figure 3 Impact on annual softwood log trade flows (million m3) as a result of restricting New Zealand's softwood log production in 2016 and 2021 for (a) Canada Exports, (b) US Exports, (c) Europe Exports, (d) Russian Federation Exports, (e) New Zealand Exports, (f) Chile Exports and (g) Other Exports (rest of world (ROW)). The import regions are on the x-axis. #### Combination of the proposed policy in Russia and New Zealand Table 7 presents the combination effects of reducing the Russian softwood log export tax and restricting New Zealand softwood log production on annual production, consumption, trade and prices in 2016 and 2021. The model projected that the softwood log prices in Russia and New Zealand would increase by$14.7/m3 and decrease by $2.5/m3 in 2021, respectively, in response to the proposed policy changes in these two countries. Because of the price changes, Russian softwood log production would increase significantly by 22.2 million m3, whereas New Zealand's timber harvest would also significantly decrease by 5.6 million m3 in 2021. In addition, considerable reallocation of trade flows was observed among China, Russia and New Zealand in the global softwood log market in this combined scenario (Figure 4). Specifically, China would source more Russian softwood logs to meet its timber demand when New Zealand restricts its softwood log production and Russia also reduces its softwood log export tax. It is projected that the exports of Russian softwood logs to China would increase by 23.7 million m3 in response to the reduced imports of 8.1 million m3 from New Zealand in 2021. Overall, the annual total log exports from Russia would increase by 27 million m3, whereas the annual total log exports from New Zealand would decrease by 5.7 million m3. The annual total trade volume in the world would increase by 7.8 million m3 in 2021 in response to the combined policies in Russia and New Zealand. Table 7 Impact on annual production, consumption, trade (net exports or imports) and prices as a result of the combination of reducing the Russian softwood log export tax and restricting New Zealand's softwood log production in 2016 and 2021a Production (million m3Consumption (million m3Trade (million m3Price (US$)
2016
Export
US −2.36 0.44 −2.80 −2.74
Europe −4.30 1.20 −5.50 −2.07
Russia 21.41 −4.38 25.78 13.81
New Zealand −2.90 0.11 −3.01 −2.74
Chile −1.44 0.36 −1.80 −2.74
Rest of world −2.57 0.64 −3.22 −2.74
Import
US −0.64 0.15 0.79 −0.33
Japan −0.16 0.10 0.26 −2.74
Korea −0.05 0.04 0.09 −2.74
China −1.63 0.66 2.29 −2.74
Europe −3.42 0.96 4.38 −2.74
Rest of world −1.26 0.39 1.65 −2.74
Total 0.67 0.67 9.45 n/a
2021
Export
US −2.09 0.43 −2.52 −2.42
Europe −5.10 1.45 −6.55 −2.35
Russia 22.18 −4.86 27.04 14.66
New Zealand −5.57 0.10 −5.66 −2.46
Chile −1.31 0.31 −1.62 −2.22
Rest of world −2.30 0.60 −2.90 −2.31
Import
US 0.00 0.00 0.00 0.03
Japan −0.15 0.08 0.23 −2.35
Korea −0.04 0.04 0.08 −2.37
China −1.45 0.61 2.06 −2.26
Europe −3.05 0.88 3.94 −2.31
Rest of world −1.14 0.35 1.49 −2.22
Total −0.01 −0.01 7.80 n/a
Production (million m3Consumption (million m3Trade (million m3Price (US$) 2016 Export Canada 0.00 0.00 0.00 0.00 US −2.36 0.44 −2.80 −2.74 Europe −4.30 1.20 −5.50 −2.07 Russia 21.41 −4.38 25.78 13.81 New Zealand −2.90 0.11 −3.01 −2.74 Chile −1.44 0.36 −1.80 −2.74 Rest of world −2.57 0.64 −3.22 −2.74 Import US −0.64 0.15 0.79 −0.33 Japan −0.16 0.10 0.26 −2.74 Korea −0.05 0.04 0.09 −2.74 China −1.63 0.66 2.29 −2.74 Europe −3.42 0.96 4.38 −2.74 Rest of world −1.26 0.39 1.65 −2.74 Total 0.67 0.67 9.45 n/a 2021 Export Canada 0.00 0.00 0.00 0.00 US −2.09 0.43 −2.52 −2.42 Europe −5.10 1.45 −6.55 −2.35 Russia 22.18 −4.86 27.04 14.66 New Zealand −5.57 0.10 −5.66 −2.46 Chile −1.31 0.31 −1.62 −2.22 Rest of world −2.30 0.60 −2.90 −2.31 Import US 0.00 0.00 0.00 0.03 Japan −0.15 0.08 0.23 −2.35 Korea −0.04 0.04 0.08 −2.37 China −1.45 0.61 2.06 −2.26 Europe −3.05 0.88 3.94 −2.31 Rest of world −1.14 0.35 1.49 −2.22 Total −0.01 −0.01 7.80 n/a aThis scenario simulated the combination of reducing the Russian softwood log export tax to 8% in 2015 following World Trade Organization accession and also restricting New Zealand's softwood log production in plantation forests with a projected 0% annual timber supply growth rate from 2011 to 2021 due to social and environmental considerations. Figure 4 Impact on annual softwood log trade flows (million m3) as a result of the combination of reducing the Russian log export tax following World Trade Organization accession and restricting New Zealand's softwood log production in 2016 and 2021 for (a) Canada Exports, (b) US Exports, (c) Europe Exports, (d) Russian Federation Exports, (e) New Zealand Exports, (f) Chile Exports and (g) Other Exports (rest of world (ROW)). The import regions are on the x-axis. Figure 4 Impact on annual softwood log trade flows (million m3) as a result of the combination of reducing the Russian log export tax following World Trade Organization accession and restricting New Zealand's softwood log production in 2016 and 2021 for (a) Canada Exports, (b) US Exports, (c) Europe Exports, (d) Russian Federation Exports, (e) New Zealand Exports, (f) Chile Exports and (g) Other Exports (rest of world (ROW)). The import regions are on the x-axis. ## Discussion and conclusions The objective of this study is to investigate future global softwood log trade flows and also to evaluate the potential effects on the world softwood log market of three what if scenarios involving reduced Russian softwood log export taxes, restricted growth in softwood log production in New Zealand, and a combination of policy changes in Russia and New Zealand. A recursive dynamic world spatial partial equilibrium model was first used to project the baseline softwood log supply, demand, prices, total trade and bilateral trade flows through 2021; the baseline results were then compared with the results from the three alternative scenarios incorporating the proposed policy changes in Russia and/or New Zealand. First, the results of our baseline projection revealed that China would continue to be the largest importer of softwood logs in the world; however, there would be a significant change in the major suppliers of softwood logs for imports over the 2012–2021 period. The model forecast that Canada would completely halt its log exports (representing a 100 per cent decline) based on the effects of the mountain pine beetle infestation on the timber supply in western Canada. Because most of the harvested softwood logs would enter into the domestic wood processing sectors (e.g. lumber) to meet the demand from the recovering American housing market, there would be no excess Canadian logs available for export. This result is consistent with Chang and Gaston (2014), who found that exports of US dimension lumber from Canada to the US would increase over the period of 2011–2021, despite the timber supply constraints imposed by the mountain pine beetle infestation in western Canada. In addition to the decrease in log exports from Canada, annual exports to China from the US export region and from Russia would also significantly decrease by 47 and 41 per cent, respectively, over the same period due to increases in domestic timber demand. To meet the growing domestic timber demand and to compensate for the significant decline in imports from Russia, Canada and the US, China would source its imports from the European export region (most likely from the Baltic countries – Estonia, Latvia and Lithuania – according to the actual trade flow data from recent years), New Zealand, Chile and the rest of world export region (with annual imports from those regions projected to significantly increase by 149, 49, 597 and 50 per cent, respectively). Although we did not specifically consider the substitutability of logs from different export regions in the model (i.e. cross-price elasticities for softwood logs) and assumed countries/regions to trade homogeneous softwood logs, the above results are consistent with Sun (2014), who used a Rotterdam demand system to assess China's roundwood import demand by supplying source and product type and found that there is little competition among softwood log-supplying countries in the Chinese timber market. The results of the scenario analysis showed that reducing Russian ad valorem softwood log export taxes would increase Russian log production, exports and prices over the 2012–2021 period. These results are consistent with Turner et al. (2008), Solberg et al. (2010) and van Kooten and Johnston (2014), who found that liberalizing Russian export taxes would increase the Russian log harvest, exports and prices, whereas the opposite effects were observed following the imposition of the log export tax. The results of our bilateral trade flow analysis also suggested that the largest beneficiaries of Russian log trade liberalization would be China (in terms of volume) and the European import regions (in terms of percentage) due to geographic proximity, with projected annual imports from Russia in those regions increasing dramatically by 263 per cent (from 8.1 to 29.5 million m3) and by 640 per cent (from 0.3 to 2.0 million m3) in 2021, respectively. In addition, the significant increases in total exports from Russia and total trade in the world may also support the findings of Mäkelä (2009), who used export data to assess the competitiveness of Russian wood products and observed that the highest revealed comparative advantage of Russian wood products are products that required little processing, such as untreated softwood logs, due to its significant market share with respect to these products. The simulation results show that reducing the Russian softwood log export tax would lead to only modest reductions in its softwood log consumption over the projection period (e.g. a decline of 5 per cent (4.4 million m3) in 2021) might also imply that imposing the softwood log export tax would not be an effective strategy to achieve the goal of encouraging forest industry development. Indeed, strategies such as improving the investment climate in Russia would have a greater impact on the development of the forest sector than imposing a log export tax as suggested by Solberg et al. (2010). The simulated results of restricted log production in New Zealand showed that such restrictions would significantly reduce log exports to the world (primarily to China). It is projected that exports to China from New Zealand would drop by 29 per cent (4.8 million m3) in 2021 under this scenario compared with the baseline projection. However, this decline in exports would be largely offset by increased exports from other export regions, such as Russia and the European export region, during the same period. Thus, restricting New Zealand's log production would cause global prices to marginally increase by$0.5/m3 in 2021 and would not substantially affect production, consumption or total trade in other regions of the world.

Overall, our results indicate that reducing Russian softwood log export taxes would cause an increase in global softwood log trade, whereas the restriction of softwood log production in New Zealand would lead to a decrease in global softwood log trade. When further combining the proposed policies in Russia and New Zealand, we found that Russia would export more softwood logs to the world in response to the timber harvest constraint in New Zealand in addition to reduced Russian softwood log export taxes. The impact on total trade and prices as a result of reducing Russian softwood log export taxes would be greater than the impact of restricting New Zealand timber production due to the comparatively large log production capacity in the Russian primary forestry and logging sector, therefore having larger effects on the global softwood log market. In addition, under any scenario, China would significantly adjust its softwood log imports from different export regions in response to these proposed trade policy changes or market shocks in Russia and/or New Zealand. This result is also reflected in the actual trade flow data from recent years that shows that China's softwood log imports have been diversified, transitioning from one region (Russia) dominating the market to various supply regions playing a larger role due to China's strong demand for softwood logs on the world market (Global Trade Information Services Inc., 2014).

Several refinements could be made to our research to improve the accuracy of the results. First, following most spatial partial equilibrium models (e.g. Buongiorno and Zhu (2013); Kallio et al., 2004), we used a single year as the base year, and the year 2011 was the latest year of available trade flow data when we prepared the manuscript. As there were no significant global events in that year (e.g. the global financial crisis in 2008 to 2009) and due to the good calibration of the model for the base year, it is apparent that the model can give plausible projections for scenario analysis. Nevertheless, given that our model is based on recursively dynamic programming for every 5-year period to allow enough time for the global softwood log market to adjust to the proposed policy changes, it would be interesting to investigate whether the impact estimates changed when using 5-year average (e.g. 2006–2011) data for model calibration to minimize the influence of significant events or market fluctuations in any single year on the results and analyses.

Additionally, to more easily distinguish and compare the effects of alternative policy scenarios, supply/demand price elasticities and transportation costs were taken from the literature and assumed to remain constant over the projection period. Discussing the sensitivity of these assumptions is beyond the scope of this study, as they may potentially interfere with (deviate from) the changes brought about by the proposed forest policies of interest and also require more assumptions. However, it is possible that altering these assumptions could have different levels of impact on demand, supply and trade flows in the regions (Kallio et al., 2006). For example, if Chinese importers show more preference for New Zealand softwood logs than Russian softwood logs in the future, which was not internalized in our model via cross-price elasticities, some of the trade flow projections in our study may be too high. Thus, caution must be taken when interpreting our results along with the awareness of the associated assumptions behind them. Future research could investigate these factors more completely by explicitly incorporating uncertainty issues into the analysis. This includes incorporating recursive dynamic models with stochastic modelling (see Kallio, 2010) or further adapting an intertemporal optimization model (see Latta et al., 2013) to fully account for how the actors will react to the proposed policy changes when future supply/demand conditions are considered endogenously.

Finally, another way this study could be improved is by including additional forest products in the trade flow model. For instance, adding hardwood logs to the model and considering the input–output relationships of logs with different categories of lumber (e.g. appearance, construction and utility and economy grades) and biomass (e.g. wood chips and wood pellets), if the necessary data become available, could provide more valuable information. Adding these products and product categories would better help policy makers and forest industries to examine the linkage effects among all related forest products in response to changes in policy or market scenarios, including the analysis of so-called bio-pathways of resource allocation for value chain optimization. Disaggregating softwood logs into two product groups (i.e. sawlogs and pulpwood) was initially undertaken by the authors, but it was discovered that the majority of softwood logs were traded in sawlog form internationally, and no significant price variation was observed between sawlogs and pulpwood (Global Trade Information Services Inc., 2014). Thus, unlike other forest products (e.g. hardwood logs, lumber) that generally show significant price variation among product groups, species or place of production, it is plausible to consider softwood logs as a homogeneous good in a trade flow analysis. Refinements such as these are intended for future research by the authors.

Notwithstanding the above issues, we believe our analysis has provided a valuable application of spatial partial equilibrium models to the forest sector. The findings of this study can help forest managers and policy makers assess the global impact of potential changes in trade policy and supply constraints in these two important softwood log supply regions and also highlight China's role in the world softwood log market.

None declared.

## Funding

This research was funded by the Natural Sciences and Engineering Research Council of Canada (NSERC) Strategic Research Network on Value Chain Optimization.

## Acknowledgements

We thank Editor Dr Anssi Ahtikoski and one anonymous reviewer for their valuable comments and suggestions that have greatly improved the article.

## References

Abbott
B.
Stennes
B.
van Kooten
G.C.
2009
Mountain pine beetle, global markets, and the British Columbia forest economy
.
Can. J. For. Res
.
39
,
1313
1321
.
Buongiorno
J.
Zhu
S.
2013
Using the global forest products model
.
Staff Paper Series#78
.
Department of Forest and Wildlife Ecology, University of Wisconsin
.
British Columbia Ministry of Forests and Range
.
2007
Timber Supply and The Mountain Pine Beetle Infestation in British Columbia: 2007 Update
. .
British Columbia Ministry of Forests, Lands, and Natural Resource Operations
.
2012
2011/12 Annual Service Plan Report
. .
Chang
W.-Y.
Gaston
C.
2014
.
Can. J. For. Res
.
44
,
1
13
.
CIBC World Markets
.
2007
Russia Plans to Dramatically Increase Its Export Tax on Logs (Equity Research Industry Update).
.
Food and Agriculture Organization of the United Nations (FAO)
.
2009
State of the Worlds Forest
. .
Food and Agriculture Organization of the United Nations (FAO)
.
2014
FAOSTAT-Forestry.

.
Gaston
C.
Marinescu
M.
2006
.
Research Paper W-2270
.
.
Gaston
C.
Delcout
G.
Cohen
D.
1999
Japan's value added market: wood product attributes and competition. Competitor analysis two
.
Research Paper 1664-2
.
.
2014
.
http://www.gtis.com/english/GTIS_GTA.html (accessed on 10 February, 2015)
.
Houck
J.P.
1986
.
Macmillan
,
191
.
Kallio
A.M.
Moiseyev
A.
Solberg
B.
2004
The global forest sector model EFI-GTM-the model structure
.
EFI Technical Report 15
.
European Forest Institute
.
Kallio
A.M.
Moiseyev
A.
Solberg
B.
,
2006
Economic impacts of increased forest conservation in Europe: a forest sector model analysis
.
Environ. Sci. Policy

9
,
457
465
.
Kallio
A.M.
2010
Accounting for uncertainty in a forest sector model using Monte Carlo simulation
.
For. Policy Econ
.
12
,
9
16
.
Latta
G.S.
Sjølie
H.K.
Solberg
G.
2013
A review of recent developments and applications of partial equilibrium models of the forest sectors
.
J. For. Econ
.
19
,
350
360
.
Lindo System Inc
.
2013
Whats Best-User’s Manual
.
Lindo System Inc
.
468
.
Mäkelä
T.
2009
The Russian Forest Industry: a Case of Competitiveness and Export Taxes
.
Master Thesis
,
Department of Economics, Helsinki School of Economics
. .
Miller
R.
Dickinson
Y.
Reid
A.
2007
Māori connections to Forestry in New Zealand
. In:
Feary
S.
(ed.),
Forestry for Indigenous Peoples: Learning from Experiences with Forest Industries
.
Australian National University
, pp.
13
22
.
Ministry of Agriculture and Forestry
.
2010
New Zealand Wood Availability Forecasts 2010-2040
. .
Ministry for Primary Industries
.
2014
Statistics & Forecasting-Forestry
. .
.
2012
.
.
Oswalt
S.
Smith
B.
Miles
P.
Pugh
S.
2014
Forest Resources of the United States, 2012
.
US Department of Agriculture
,
Forest Service
.
http://www.fia.fs.fed.us/program-features/rpa/ (accessed on 15 February, 2015)
.
Paris
Q.
Drogue
S.
Anania
G.
2011
.
Econ. Model.

28
,
2509
2516
.
Random Lengths International
.
2012
Special Report: China and Russia News & Statistics
..
Rhodes
D.
Novis
J.
2004
Impact of incentives on the development of plantation forest resources in New Zealand
. In:
Enters
T.
Durst
B.
(eds),
What Does It Take? The Role of Incentives in Forest Plantation Development in Asia and the Pacific
.
Food and Agriculture Organization of the United Nations
, pp.
151
196
.
Samuelson
P.
1952
Spatial price equilibrium and linear programming
.
Am. Econ. Rev
.
42
,
283
303
.
Sedjo
R.
1999
The potential of high-yield plantation forestry for meeting timber needs
.
New Forests

17
,
339
359
.
Simeone
J.
2012
Timber export taxes and trade between Russia and China: Development of the forestry sector in the Russian Far East
.
For. Chron.

88
,
585
592
.
Simeone
J.
Eastin
I.
2012
Russia’s Log Export Tariff and WTO Accession
.
Center for International Trade in Forest Products
.
University of Washington
.
www.cintrafor.org (accessed on 10 February, 2015)
.
Solberg
B.
Moiseyev
A.
Kallio
A.M.
Toppinen
A.
2010
Forest sector market impacts of changed roundwood export tariffs and investment climate in Russia
.
For. Policy Econ
.
12
,
17
23
.
Sun
C.
2014
Recent Growth in China's roundwood import and its global implications
.
For. Policy Econ
.
39
,
43
53
.
Takayama
T.
Judge
G.
1971
Spatial and temporal price and allocation models
.
North-Holland Publishing Co.
,
528
pp.
Turner
J.
Buongiorno
J.
Katz
A.
Zhu
S.
2008
Implications of the Russian roundwood export tax for the Russian and global wood products sectors
.
Scandinavian J. For. Res
.
23
,
154
166
.
van Kooten
G.C.
Johnston
C.
2014
Global impacts of Russian log export restrictions and the Canada-U.S. lumber dispute: Modeling trade in logs and lumber
.
For. Policy Econ
.
39
,
54
66
.

### Appendix

Appendix A. Derivation of net demand and supply functions

Following Chang and Gaston (2014), suppose net demand region i (i = 1 … n) and net supply region j (j = 1…m) have the following continuous linear domestic demand and supply functions for softwood logs:

(A1)
$qid=ai−biρi$

(A2)
$qis=ci+diρi$

(A3)
$qjd=aj−bjρj$

(A4)
$qjs=cj+djρj$
where, for net demand region i, ai and bi are the intercept and slope, respectively, of the domestic demand function; ci and di are the intercept and slope, respectively, of the domestic supply function; $qid$ and $qis$ are the domestic consumption and production, respectively; and ρi represents the domestic market price. For net supply region j, aj and bj are the intercept and slope, respectively, of the domestic demand function; cj and dj are the intercept and slope, respectively, of the domestic supply function; $qjd$ and $qjs$ are the domestic consumption and production, respectively; ρj is the domestic market price.

The intercepts and slopes of domestic demand and supply functions for each region could be derived from the base year data of production, consumption, average market prices and the own-price elasticities of domestic demand and supply, which are presented in equations (A5–A12) below:

(A5)
$bi=eidqidρi$

(A6)
$ai=qid+biρi$

(A7)
$di=eisqisρi$

(A8)
$ci=qis−diρi$

(A9)
$bj=ejd(qjdρj)$

(A10)
$aj=qjd+bjρj$

(A11)
$dj=ejsqjsρj$

(A12)
$cj=qjs−djρj$
where $eid,eis$ and $ejd,ejs$ represent the own-price elasticities of domestic demand and supply for softwood logs in region i and region j, respectively. To derive the intercept and slope parameters of net demand and supply functions for softwood logs in equations (1) and (2) (i.e. αi, βi, γj, δj), net demand and supply elasticities are required and are presented in equations (A13) and (A14) below:
(A13)
$ϵid=eidqidMi−eisqisMi$

(A14)
$ϵjs=ejsqjsXj−ejdqjdXj$
where $ϵid$ is the net demand elasticity of softwood logs in region i and $ϵjs$ is the net supply elasticity of softwood logs in region j. Mi is the import quantity of softwood logs in region i, and Xj is the export quantity of softwood logs in region j, which are from the Global Trade Atlas®. The intercepts and slopes of the net demand and net supply functions can therefore be derived in equations (A15)–(A18) below:
(A15)
$αi=(ci−ai)(bi−di)$

(A16)
$βi=1ϵidMiρi$

(A17)
$γj=(cj−aj)(bj−dj)$

(A18)
$δj=1ϵjsXjρj$

Appendix B. Projections of future domestic consumption and production

Suppose region i (i = 1, …, n) has the following domestic demand and supply functions for softwood logs in time period t:

(A19)
$qitd=ait−bitpit$

(A20)
$qits=cit+ditpit$
where ait and bit are the intercept and slope, respectively, of the domestic demand function of region i in time period t; cit and dit are the intercept and slope, respectively, of the domestic supply function of region i in time period t; $qitd$ and $qits$ are the domestic consumption and production, respectively, of region i in time period t; and pit represents the domestic market price of region i in time period t. We assumed that changes in consumption and production for each time period were measured by changing the intercepts of the domestic demand and supply functions and that each time period was dependent on the changes in previous time periods. Specifically, the intercepts of the domestic demand and supply functions in equations (A19) and (A20) for the next time period are defined as follows:
(A21)
$ai(t+1)=ait+qitd((RiD)((t+1)−t))$

(A22)
$ci(t+1)=cit+qits((RiS)((t+1)−t))$
where t + 1 is the year of the next period (e.g. 2016) and t is the year of the current period (e.g. 2011). $RiD$ and $RiS$ are the expected annual demand and supply changes (per cent) for softwood logs in region i, respectively. Because the intercept parameters of the domestic demand and supply functions were updated for each 5-year period, the intercept parameters of the net demand and supply functions in the model were also changed accordingly.

Appendix C. Changes in net supply (export) function for the ad valorem export tax

Following Houck (1986), suppose region j (e.g. Russia) has the following domestic demand and supply functions of softwood logs:

(A23)
$qjd=aj−bjpj$

(A24)
$qjs=cj+djpj$

The net supply function in region j can be expressed as follows

(A25)
$qjs−qjd=Xj=(cj−aj)+(dj+bj)pj$
where Xj is the excess supply (export) in region j. The inverse net supply function is then derived as
(A26)
$pj=−(cj−aj)(dj+bj)+1(dj+bj)Xj$
Before imposing the ad valorem export tax, the domestic price in region j is the same as the world price (assuming that the cost of moving logs between regions is zero):
(A27)
$pj=pjw$

After the ad valorem export tax t, which creates a wedge between world and domestic prices, the world price is domestic price plus export tax as shown below:

(A28)
$pjw=pj(1+t)orpj=pjw(1+t)$

By substituting the new world price of equation (A28) into the net supply function of equation (A25), the new net supply and inverse net supply functions after the ad valorem export tax can then be given in equations (A29) and (A30) below:

(A29)
$Xj=(cj−aj)+(dj+bj)pjw(1+t)$

(A30)
$pjw=−(cj−aj)(dj+bj)(1+t)+(1+t)(dj+bj)Xj$

Comparing the intercepts and slopes of inverse net supply functions (Equations A26 and A30) before and after the export tax, both the price intercept and slope of Equation (A30) increase by (1 + t) after the ad valorem tax. However, the quantity intercept remains the same at (cjaj) before and after the ad valorem tax (when Pj = 0), which indicates that imposing an ad valorem tax will pivot the net supply curve inwards from the same point on the horizontal quantity axis.