Landowners Are Interested in Payment for the Ecosystem Services of Forestry: The Case of Korean Private Forests

Abstract

As no market or compensation scheme exists for the provision of ecosystem services for private forest owners, landowners have become more interested in timber and forest production or the conversion of forests into other lands. We tested a hypothesis on the willingness of landowners to accept payments for the ecosystem services of nature-friendly forestry. The conditional logit model analysis showed that restoring planted forests into natural forests and the extension of rotation ages negatively impact the utility of forest owners, who thus require financial compensation for such activities required for higher levels of public benefits. Forest owners are willing to participate in the conversion of artificial commercial forests into natural forests if they are compensated for as much as KRW 185,280 and 212,060/ha/year for extending the rotation age by 50 years, whereas no compensation is demanded for restoring half the land into natural forests, forest protection activities, and thinning. We suggest that the payment for an ecosystem services scheme should be designed based on the detailed information on forestry operations that enhance public goods and ecosystem services. Our results provide a basis for governments to design compensation protocols and the level of compensation for ecosystem services through the management of private forests.

1. Introduction

1.1. Research Background

To prevent and respond to irreversible changes in the ecosystem, international society has established and achieved common goals to realize sustainable development and live harmoniously with the ecosystem. According to the Millennium Ecosystem Assessment [1], over the past 50 years, humans have changed ecosystems more rapidly and extensively than at any other time in history, and these ecosystem services have deteriorated as humans continue to engage in social and economic activities without considering the real value of the ecosystem. Market failure on externalities and public goods showed that forest owners are more interested in timber production or land-use change. Payment for ecosystem services (PES) is a representative economic policy tool that allows providers to voluntarily supply services through financial incentives [2]. PES internalizes the externalities by motivating landowners to take actions for social and environmental benefits by providing financial incentives, even though the private benefits may be limited [3], and pay some incentives. Many countries have implemented PES programs for natural resource conservation and various ecosystem services. Costa Rica approved the Forest Law in 1996 and created a PES program, Pago por Servicios Ambientales, to compensate landowners for reversing deforestation and forest conservation [4]. Similarly, the Forest Stewardship Program and Landowner Incentive Program in the USA helps private forest owners enhance economic and environmental benefits by protecting their forests. Moreover, the Forest Stewardship Program provides technical assistance for forest owners to achieve their multi-purpose plans [5], and the Landowner Incentive Program incentivizes private landowners to protect and restore habitats for at-risk plants and animal species [6]. The Natural Values Trading in Finland and Nature Conservation Agreements in Sweden also support the conservation of biodiversity and habitats in private forests.

As participation in PES is based on the willingness of landowners, it is important to understand the factors and motivations affecting their participation, leading to improved program design and enrollment [7,8,9]. Many studies have contributed to the design of effective PES schemes by identifying the preferences and willingness of landowners to ensure participation; for example, demographic and socioeconomic factors such as gender, age, and education in Kenya, Brazil, and Ethiopia [10,11,12], and the characteristics of their lands, such as the size and influence participation by landowners [13,14,15,16] in PES schemes. Jones et al. (2020) argued that participation can be determined by both financial and nonfinancial motivators [9]. When the financial benefit is the driver for participation, the compensation amount should be higher than the opportunity cost of using the land for other activities, such as timber and farming, and this opportunity cost might differ by region and community [9]. Studies have found a willingness to accept (WTA) when landowners are willing to participate in the program by suggesting forestry or farming activities for the conservation and provision of public ecosystem services [16,17]. Additionally, social and environmental reasons, rather than monetary motivation, can be one of the determinants of participation by landowners. Here, the factors can be landowners’ intentions, behaviors, attitudes, subjective norms, and individuals’ environmental values [18,19]. Moreover, management preferences have an impact on participation. Kline et al. (2000) asked forest owners if they were willing to forego harvesting with some tax reduction within 200 feet of a riparian area for the wildlife habitat [13]. Matta et al. (2009) examined the preferred management practices of forest owners in a conservation incentive program [20], and Tyrväinen et al. (2021) identified the preferences of forestry attributes, such as harvesting restrictions and the length of outdoor routes [21]. Godoy et al. (2022) investigated landowners’ preferences for conservation incentives for activities allowed on their lands and annual payments [22]. Similarly, while there is a growing interest in PES, the heterogeneity for participating in PES varies by country and community. Therefore, a case study of various countries is required to investigate the factors affecting voluntary participation and to develop an effective PES scheme for sustainable ecosystem services.

To develop an efficient scheme and ensure the sustainable use of ecosystem services, the supply of ecosystem services should adhere to the demands of society [23]. Hence, it is essential to ensure the supply of ecosystem services by creating a forestry incentive program with a subscription of detailed forestry operations geared toward environmental services, such as carbon sequestration, landscape preservation, cultural services, and the amount of compensation comparable to the opportunity costs or value of the services. Therefore, this study aimed to provide information for designing PES schemes and to estimate the amount of compensation to be paid to private forest owners by identifying their preferences for forestry operations to provide additional ecosystem services. We addressed the following aspects in this paper: (1) the type of forestry operations for which the forest owner must be compensated for their contributions toward providing additional forest ecosystem services; and (2) the willingness of forest owners to accept their efforts geared toward additional provisions of ecosystem services from their forest.

1.2. Private Forest and Payment for Ecosystem Services Scheme in the Republic of Korea

Forests in the Republic of Korea (ROK) cover an area of 6.3 million ha, accounting for approximately 63% of the total land area. Here, private forests owned by individuals, firms, clans, temples, and private schools accounted for 66.1%, followed by national forests belonging to the state (26.2%) and public forests belonging to counties and provinces (7.7%). Thus, private forests in South Korea are expected to play a critical role in providing ecosystem services to society. However, owing to privatization during Japanese colonial imperialism and the demand for forest lands as graveyards of families, private forest lands have been fragmented. Consequently, the size of forests owned by 97% of private landowners is less than 10 ha, with an average of 2 ha per forest owner [24]. As 67% of landowners own less than 1 ha, under such a small forest structure, it is unlikely for private forest owners to run forestry for deriving profits, which discourages forest management.

The government of the ROK has attempted to revive private forest management with a series of fiscal and informational measures, such as the promotion of forestry cooperatives, forest management contract schemes, and multiple forest management, but to no avail [25]. Fiscal policies, such as PES and direct payments for private forest owners, were introduced to promote biodiversity conservation and the provision of ecosystem services. The Act on the Conservation and Use of Biological Diversity was legislated in 2021, and PES is now being implemented under this law. The Ministry of Environment contracts with the owner of land designated as a protected area and pays the amount of costs or losses to owners or custodians of the land for conservation activities that can enhance ecosystem services. For example, feeding or providing habitats for wildlife, improving water and air quality, disaster prevention, and maintaining natural landscapes in their lands are paid on a contract basis [26]. Although the PES scheme under the Biodiversity Conservation Act is intended to internalize the positive externalities of land management in protected areas, the target has been limited to protected areas. In contrast, Korea Forest Service (KFS) introduced a direct payment scheme for forest owners to not only stabilize the income of owners but also to promote the provision of public forest ecosystem services in 2021. This scheme sets a payment scheme for forest owners who engage in forestry operations, including silviculture and the production of forest products. However, there is a limitation to enhancing environmental services because the basis of payment is the proof of forestry operations and not the increase in ecosystem services supply. Subsidies are granted to owners who meet the conditions based on their land size, without considering the ecosystem service output. Thus, a constraint still exists in enlarging environmental services to the public through forestry with current schemes that have not established a comprehensive standard for the increase in the ecosystem services of forestry activities. Additionally, research on identifying private owners’ willingness to participate in such schemes and forestry operations to provide ecosystem services is scarce in the ROK.

2. Materials and Methods

2.1. The Model

The stated preference method has been broadly applied to estimate non-use values in many studies [27]. The choice experiment (CE) method used in this study suggests alternatives comprising various attributes and levels to respondents and estimates the size of utility for the selected attribute, facilitating the comparison of preferences for each attribute [28,29,30,31]. CE is theoretically based on the random utility model [32], multi-attribute utility theory [33], and discrete choice model [34,35]. The random utility model is based on the random utility theory, where an individual (n) chooses one alternative (i) from a number of choice sets ($C_n$), and the indirect utility of the selected alternative (i) is greater than that of any other option (j). It comprises an observable deterministic part ($V_{ni}$) and an unobservable stochastic part (${\epsilon }_{ni}$).

$$U_{ni}=V_{ni}\left(x_i,t\right)+{\epsilon }_{ni}$$

$$U_{ni}>U_{nj}=>V_{ni}+{\epsilon }_{ni}>V_{nj}+{\epsilon }_{nj} \forall j \ne i ;i,j \in C_n$$

Here, $x_i$ represents the characteristics of goods or services, namely, in this study, the outputs are contingent on forestry operations; and $t$ represents the compensation amount (WTA). We assumed that respondents would choose option i when the utility of goods and services produced by the combination of forestry operations is greater than any other alternative.

According to multiple-attribute utility theory, a commodity can be represented by several attributes [33]. In this theory, consumers obtain utility from these attributes of goods and services rather than the forestry operation per se and choose the alternative that can provide the greatest utility from the chosen forestry option. The discrete choice model assumes that consumers take actions to maximize utility. If the utility of selecting option i is greater than that of selecting option j, option i is selected. The probability of selection is expressed as follows:

$$P{r}_{ni}=Pr\left(V_{ni}+{\epsilon }_{ni}>V_{nj}+{\epsilon }_{nj}\right)$$

In this study, the utility of forest owners is assumed to be determined by profit ($M_i)$. This profit is the value of total revenue (${\mathsf{\Pi }}_i)$ minus the net of the total cost of operating forestry option i ($C_i)$. According to the dual theory, the forest owner’s utility is maximized when the profit of forestry operations is maximized or the cost of practicing the forestry option is minimized. At this time, forest owners’ revenue can be obtained from the ecosystem services available from forestry option i; this is multiplied by the price of ecosystem services ($P_i)$ and the quantity of goods and services ($Y_i)$ as follows.

$$M_i\left(profit\right)={\Pi }_i\left(revenue\right)-C_i\left(cost\right)={\displaystyle \sum }_{i=1}^m{P}_i{Y}_i-C_i\left(x\right)$$

However, as ecosystem services other than market goods, such as timber and edible wild fruits and vegetables, lack markets in the ROK, we assumed that the forest owner can obtain the compensation amount ($C{S}_i)$ for non-market ecosystem services. Here, market goods are denoted as $i=1$, whereas the others ($i\ge 2$) have no market value, and the non-market value is expressed as the compensation amount ($C{S}_i)$. The cost ($C_i)$ is the cost of forestry operations for supplying ecosystem services. The value or utility of forestry operation $x_i$ for the forest owner can be expressed as follows:

$${\displaystyle \sum }_{i=2}^m{P}_i{Y}_i=0$$

$$Utility of x_i=M_i\left(x, P\right)=P_1{Y}_1+{\displaystyle \sum }_{i=2}^{m}C{S}_i-C_i\left(x\right)$$

The CE method has the advantage of deriving the relative importance by calculating the marginal WTA (MWTA) through the comparison of monetary and non-monetary values [36]. The MWTA shows the compensation amount for accepting an additional unit of forest attributes, and the non-monetary attribute can be monetized as an implicit value. When the indirect utility function is linearly specified, MTWA can be represented as ${\alpha }_k$/${\alpha }_p$, where ${\alpha }_k$ is the coefficient of the forestry operation attributes and ${\alpha }_p$ is the coefficient of the monetary compensation attribute.

The change in the probability that forest owners n choose alternative i given a change in one unit of the attribute $z_{ni}$ with other alternative constants can be expressed through the marginal effect [37].

$$\frac{\partial P_{ni}}{\partial z_{ni}}=\frac{\partial [\left({\epsilon }^{V_{ni}}\right)/({{\displaystyle \sum }}_j{\epsilon }^{V_{nj}})]}{\partial z_{ni}}$$

$$=[{\epsilon }^{V_{ni}}/\sum {\epsilon }^{V_{nj}}\left]\left(\frac{\partial V_{ni}}{\partial z_{ni}}\right)-\right[{\epsilon }^{V_{ni}}/{(\sum {\epsilon }^{V_{nj}})}^2\left({\epsilon }^{V_{ni}}\right)\left(\frac{\partial V_{ni}}{\partial z_{ni}}\right)$$

$$=\frac{\partial V_{ni}}{\partial z_{ni}}\left(P_{ni}-P_{ni}{}^2\right)$$

$$=\frac{\partial V_{ni}}{\partial z_{ni}}{P}_{ni}\left(1-P_{ni}\right)$$

In this study, the utility of forest owners from the forestry option providing ecosystem services was maximized by selecting the most preferred option among the three alternatives combined with different forestry options and compensation levels, and this value can be efficiently estimated through the conditional logit model. When the observable deterministic part ($V_{ni}$) is assumed to be linear, and an unobservable stochastic part (${ϵ}_{ni}$) follows irrelevant independence alternatives (IIA) and Gumbel distribution, the conditional logit probability model can be expressed as follows:

$$P{r}_{ni}=\frac{EXP\left(V_{nj}\right)}{{{\displaystyle \sum }}_{j=1}^{J}EXP\left(V_{nj}\right)}$$

2.2. Data Collection

2.2.1. Design

In this study, prior to setting the choice set, the analytic hierarchy process (AHP) was applied to determine the attributes and levels. AHP is a decision-making method that provides optimal alternatives by analyzing the importance of each attribute in a hierarchical structure [38]. In this study, four forestry activity attributes and the levels of each attribute to promote environmental services were determined by 15 experts from the Korean Forest Economic Association. The upper limit of the amount of compensation (WTA) was set based on the sum of the costs of each activity, assuming that each activity is independent. The following levels of WTA were determined by decreasing the amount by KRW 150,000 (Korean Won) from the upper limit of KRW 400,000 (

Table 1. Options of forestry operations (x) and willingness to accept corresponding to forestry operations.

Options of Forestry Operations and Compensation for Conservation		Levels
Restoration to natural forest from planted forest	Induce ecological transition and implement appropriate thinning for the ecological restoration of natural forests. Artificial plantation after harvesting is not allowed.	No restoration
		50% restoration
		100% restoration
Extension of timber harvesting rotation age	Extend the rotation age from the standard rotation age under law “Creation and Management of Forest Resources Act.”	No extension
		Extension of 20 years
		Extension of 50 years
Forest protection activity (Monitoring, Waste cleaning, etc.)	Protect the forest by monitoring or conducting regular cleanup activity.	No activity
		Once a month
		Twice a month
Thinning	Conduct timely thinning and tending of young trees to improve forest quality.	No thinning
		Several times every more than 5 years
WTA	Willingness to accept the forestry activities for the provision of forest ecosystem services.	150,000 KRW/ha/year
		300,000 KRW/ha/year
		450,000 KRW/ha/year

).

Man-made forests, which are artificial forests combined with natural vegetation, cause heterogeneous landscapes, and coniferous forests with a single species degrade biodiversity owing to allelopathic substances [39]. Conversely, natural forests provide public benefits, such as erosion and flood control, improved water quality, and habitats for wild animals in addition to the economic value for forest owners [40]. Extending the rotation age can ensure high-quality timber production as well as enhanced forest ecosystem services, particularly related to biodiversity and watershed conservation. When the rotation age was extended from 50 to 70 years, the effect of increasing the water resource storage increased by 76 m³/ha in coniferous forests and by 140 m³/ha in broadleaf forests [41]. It also has a positive effect on recreational value [42] and may increase habitats for microorganisms and wildlife in old trees [43,44]. Forest monitoring activities by forest owners enable the early detection and extinction of wildfires. Additionally, forest cleaning activities can prevent landscape damage, water contamination, and forest fires by removing waste or garbage that may cause combustion. Thinning is a forestry activity aimed at achieving sustainability and optimizing public and socioeconomic benefits. Forest stands with thinning removed more carbon than those without thinning until the final yield [45]. Thinning can gradually decrease the risk of landslides over time unless the stand density changes abruptly [46] and that of forest fires [47]. Thus, given these expected effects, we suggest four forestry operations wherein forest owners can provide additional ecosystem services through activities and levels referred to in the previous literature.

Considering all attribute levels, a total of 162 (3 × 3 × 3 × 2 × 3) choice sets are available. Consequently, we used an orthogonal design and identified a total of 16 alternative choice sets comprising forestry operations geared toward conservation contingent on compensation. However, we removed one choice set that provided compensation but had no forestry activity, thus resulting in a total of 15 choice sets. Then, we chose one card randomly among 15 sets and another random card among 14 sets. Therefore, one choice set had three options, including the option of not participating in the PES scheme (options A, B, and C) (

Table 2. Examples of choice sets.

Option A □		Option B □		Option C □
Restoration to natural forest from planted forest	No restoration	Restoration to natural forest from artificial plantations	50% restoration	Not participating in any option
Extension of timber harvesting rotation age	No extension	Extension of timber harvesting rotation age	Extension of 20 years
Forest protection activity (Monitoring, Waste cleaning, etc.)	Twice a month	Forest protection activity (Monitoring, Waste cleaning, etc.)	Twice a month
Thinning	Several times every more than 5 years	Thinning	No thinning
WTA	300,000 KRW/ha/year	WTA	150,000 KRW/ha/year

) (See Supplementary Materials for questionnaire).

By applying options of forestry operations chosen by the forest owner, the utility to them can be expressed by the following linear function:

$$\begin{gathered} V_{ni}={\beta }_{CON}·CONVERSIO{N}_{ni}+{\beta }_{ROT}·ROTATIO{N}_{ni}+{\beta }_{PRO}·PROTECTIO{N}_{ni} \\ +{\beta }_{THI}·THINNIN{G}_{ni}+{\beta }_{WTA}·WT{A}_{ni} \end{gathered}$$

The questionnaire was structured as follows. The first page of the questionnaire presented a background on the PES scheme and the study. The questionnaire then contained with questions regarding the respondent’s forest and property characteristics, including the forest size and type. The questions were followed to estimate the relative importance of ecosystem services. After answering the questions, four choice sets randomly selected from 16 choice sets with two forestry options and a status quo option were presented to the respondents for their choice. The final part of the questionnaire contained socioeconomic questions, such as their household and membership in the organization.

2.2.2. Response Rate

The survey was conducted from 14 May to 21 May 2021 by members of the Korea Forestry Successor Association. “Forestry successor” refers to any person who has the will and capacity to engage in forestry for the succession and development of forestry, meeting the requirements prescribed by the Ordinance of the Ministry of Agriculture, Food, and Rural Affairs (Forestry and Mountain Villages Development Promotion Act), around 18,996 as of 2020. The association of forestry successors was established for the efficient dissemination of forestry management and technology and the exchange of information among members, with approximately 3711 members as of 2020. We chose them as respondents by considering that they were actively engaged in forestry and would provide additional ecosystem services through the PES scheme. In total, 319 people were invited to participate in the questionnaire, and 102 people responded completely, corresponding to a response rate of 32%. By asking respondents four times with a set of three options and four different sets to choose alternatives per respondent, we obtained a total of 1224 observations.

3. Results

3.1. Socio-Economic and Demographic Background of the Respondents

Table 3. Socioeconomic backgrounds and demographics of the respondents.

Description	Frequency	Percentage (%)
Gender
Male	76	75
Female	26	25
Age (years)
26–45	11	11
46–65	85	83
66+	6	6
Education
Middle school level	4	4
High school level	30	29
College level	47	46
Graduate level	21	21
Annual household income
10 million KRW ≤	12	11.76
10 million–20 million KRW	9	8.82
20 million–30 million KRW	15	14.71
30 million–40 million KRW	20	19.61
40 million–50 million KRW	18	17.65
≤50 million KRW	28	27.45
Forest holding size
≤2.0 ha	25	24.5
2.1–5.0 ha	38	37.24
5.1–10.0 ha	15	14.7
10.1–50.0 ha	23	22.54
≥50.0 ha	1	0.98
Type of forest owned
Coniferous forest	26	25.49
Broad-leaved forest	28	27.45
Mixed forest	48	47.06

presents the results of the respondents’ socioeconomic and demographic backgrounds. Regarding the gender of the respondents, 28% were female, and most of them were male. The average age was 55.4 years; the oldest and youngest respondents were 72 and 29 years old, indicating a large difference in age. In terms of the educational background, university attendance and graduation accounted for 67%, followed by high school and graduate school graduation at 29 and 4%, respectively. As for the annual income, 28 respondents had an annual income of KRW 50 million or more, accounting for about 27% of all respondents, and respondents with annual incomes between 10 million and 20 million KRW had the least share. Although the average household income in Korea is 37.5 million KRW [24], many owners had higher incomes in this study. The respondents who owned forest areas between 2.1 and 5 ha and those who owned mixed forests accounted for the majority (47%).

3.2. Estimation of Parameters of the Conditional Logit Model

In this study, R package and c-logit code were used to estimate the parameters of the conditional logit model and Stata for identifying the heterogeneity of choice (See Supplementary Materials). The WTA was analyzed as a continuous variable, and the remaining attribute variables were analyzed as categorical variables. All estimated parameters were found to be significant within the significance levels of 10, 5, and 1%, except for activities related to extending the rotation age by 20 years and conducting forest protection activity once a month. Restoring planted forests to natural ones by 50% was a positive (+) parameter compared to no conversion but was a negative (−) parameter regarding 100% conversion. The activity of extending the rotation age by 20 years as well as conducting forest protection activities once a month did not show a significant effect on the utility of the owners; however, extending the rotation age by 50 years and conducting protection activity twice a month presented significant negative and positive effects, respectively, compared to the reference. The forest owner showed a positive response for conducting forest thinning, and this activity was also highly significant (p < 0.001) on the WTA. A summary of the analysis results is presented in

Table 4. Estimated parameters of the conditional logit model.

Variables	Coefficients	Standard Error	p-Value
50% Conversion (Reference: no restoration)	0.4364728 ***	0.1602726	0.006
100% Conversion (Reference: no restoration)	−0.552462 **	0.2193794	0.012
Extension of 20 years (Reference: no extension)	−0.2742752	0.2057582	0.183
Extension of 50 years (Reference: no extension)	−0.6323269 ***	0.2281028	0.006
Protection activity 1/month (Reference: no activity)	0.2902857	0.2303773	0.208
Protection activity 2/month (Reference: no activity)	0.4313824 **	0.1803352	0.017
Thinning (Reference: no thinning)	0.7209933 ***	0.1735843	0.000
WTA	0.0298177 ***	0.0063107	0.000

p < 0.01 ***, p < 0.05 **.

.

As the coefficient was not significant for activities related to extending the rotation age by 20 years and conducting forest protection activity once a month, the MWTA for these two operations was not estimated. The negative value of coefficients for three forestry operations of “50% restoration of planted forests into natural forests,” “forest protection activities two times a month,” and “thinning operations” imply that forest owners may enjoy utility from such operations and could engage in such operations even if they had to bear additional costs by themselves. Conversely, forest activities that have positive coefficients on MWTA, such as restoring 100% of planted forests to natural forests and extending the harvesting age by 50 years, require compensation for the opportunity cost. The forest owners demanded KRW 185,280/ha/year for the opportunity cost of 100% conversion of planted forests to natural forests and KRW 212,060/ha/year for an extension of rotation age by 50 years. Assuming that forest owners implement all significant activities, the sum of their MWTA was 135,510/ha/year (

Table 5. Estimated MWTA for conservation forestry (Unit: KRW 10,000).

Variables	MWTA	95% CI
		Minimum	Maximum
50% Conversion (Reference: no restoration)	−14.638 ***	−32.253	−3.946
100% Conversion (Reference: no restoration)	18.528 **	5.158	33.856
Extension of 20 years (Reference: no extension)	9.198	−4.761	25.495
Extension of 50 years (Reference: no extension)	21.206 ***	6.199	41.196
Protection activity 1/month (Reference: no activity)	−9.735	−30.090	5.683
Protection activity 2/month (Reference: no activity)	−14.467 **	−35.828	−2.121
Thinning (Reference: no thinning)	−24.180 ***	−53.416	−10.236

p < 0.01 ***, p < 0.05 **.

).

When a specific explanatory variable changes, the probability value, including that variable, also changes.

Table 6. Effects of marginal change in forestry operations.

Variables	dy/dx	95% CI
		Minimum	Maximum
50% Conversion (Reference: no restoration)	0.0840975 ***	0.0268649	0.1413301
100% Conversion (Reference: no restoration)	−0.1276772 **	−0.2313151	−0.0240392
Extension of 20 years (Reference: no extension)	−0.0575498	−0.1454441	0.0303444
Extension of 50 years (Reference: no extension)	−0.1408566 ***	−0.2452126	−0.0365006
Protection activity 1/month (Reference: no activity)	0.0618255	−0.0307931	0.1544442
Protection activity 2/month (Reference: no activity)	0.0892474 **	0.0204826	0.1580121
Thinning (Reference: no thinning)	0.1488774 ***	0.0825626	0.2151923
WTA	0.0063331 ***	0.0038037	0.0088624

p < 0.01 ***, p < 0.05 **.

presents the marginal effects of the explanatory variables on the probability of forestry operations. Each marginal effect was estimated relative to the baseline level (reference level). Forest owners prefer their lands being converted into natural forests by more than 8% compared to no conversion, which is in contrast to completely converting them into natural forests. The probability of choosing the option of the complete conversion of their forests into natural forests decreased by 12.8%. Similarly, compared to that of no extension of the rotation age, the probability of choosing the option of extending the rotation age by 50 years decreased by 14%. The choices of protection activity twice a month and thinning activity by forest owners increased by 8.9 and 14.9% compared to that of no action, respectively. The higher the compensation, the higher the preference for such forestry options.

To see the heterogeneity of choice, we analyzed two categorical variables, which are gender (male, female) and forest type (coniferous, broad-leaved, mixed) by dividing into groups and using the Wald test. It showed that there was difference between males and females in all forestry operations, including restoration to natural forests, extension of rotation age, forest protection, and thinning. In contrast, for the restoration of natural forests, extension of 50 years rotation age, protection activity, and thinning, there were differences in the willingness of forest owners’ participation in forestry geared to enhanced ecosystem services when the sample of respondents were grouped by forest type (

Table 7. Estimated parameters of the conditional logit model between gender.

Variables		Male	Female	a  $\mathbf{Wald} {\mathit{x}}^2$
50% Conversion (Reference: no restoration)	Coefficients	0.46749 **	0.52538	6.08 **
	Std.Err.	0.18954	0.32320
100% Conversion (Reference: no restoration)	Coefficients	−0.60768 **	−0.25982	4.75 **
	Std.Err.	0.27890	0.39336
Extension of 20 years (Reference: no extension)	Coefficients	−0.63314 **	0.46838	5.91 **
	Std.Err.	0.26046	0.36053
Extension of 50 years (Reference: no extension)	Coefficients	−1.08470 ***	0.06623	13.05 ***
	Std.Err.	0.30025	0.38425
Protection activity 1/month (Reference: no activity)	Coefficients	0.63063 **	−0.25797	4.46 **
	Std.Err.	0.29871	0.39574
Protection activity 2/month (Reference: no activity)	Coefficients	0.52756 **	0.29277	5.40 **
	Std.Err.	0.22700	0.32588
Thinning (Reference: no thinning)	Coefficients	0.91517 ***	0.38385	15.45 ***
	Std.Err.	0.23285	0.28848
WTA	Coefficients	0.02517 ***	0.03830 ***	9.21 ***
	Std.Err.	0.00829	0.01091
LR chi2(8)		140.93	46.44
Prob > chi2		0.0000	0.0000
Log Likelihood		−263.514990	−91.033388
Pseudo $R^2$		0.2110	0.2032
No. of observation		912	312

p < 0.01 ***, p < 0.05 **. ^a Wald $x^2$ value on null hypothesis (${\beta }_{male}={\beta }_{female}$).

and

Table 8. Estimated parameters of the conditional logit model between forest type owned by owners.

Variables		Coniferous	Broadleaved	Mixed	§  $\mathbf{Wald} {\mathit{x}}^2$
50% Conversion (Reference: no restoration)	Coefficients	0.48336	0.18277	0.62453 ***	6.63 ***
	Std.Err.	0.31913	0.30463	0.24249
100% Conversion (Reference: no restoration)	Coefficients	−0.04722	−0.47081	−0.77065 **	5.92 **
	Std.Err.	0.50433	0.40576	0.31677
Extension of 20 years (Reference: no extension)	Coefficients	−0.64405	−0.07391	−0.21657	0.48
	Std.Err.	0.45175	0.36905	0.31271
Extension of 50 years (Reference: no extension)	Coefficients	−0.57688	−0.54485	−0.75977 **	4.35 **
	Std.Err.	0.47462	0.39042	0.36435
Protection activity 1/month (Reference: no activity)	Coefficients	0.64910	0.32675	0.11119	0.09
	Std.Err.	0.51718	0.39392	0.36419
Protection activity 2/month (Reference: no activity)	Coefficients	0.37092	0.41552	0.55725 **	4.04 **
	Std.Err.	0.38021	0.33303	0.27730
Thinning (Reference: no thinning)	Coefficients	0.62035 *	0.57751 **	0.96911 ***	11.70 ***
	Std.Err.	0.37190	0.29520	0.28327
WTA	Coefficients	0.03609 ***	0.04080 ***	0.01600	2.59
	Std.Err.	0.01390	0.01103	0.00995
LR chi2(8)		53.41	52.48	74.49
Prob > chi2		0.0000	0.0000	0.0000
Log Likelihood		−87.550445	−96.806137	−173.69038
Pseudo $R^2$		0.2337	0.2132	0.1766
No. of observation		312	336	576

p < 0.01 ***, p < 0.05 **, p < 0.1 *. ^§ Wald $x^2$ value on null hypothesis (${\beta }_{Coniferous}={\beta }_{Broadleaved}={\beta }_{Mixed}$).

).

4. Discussion

The type and amount of ecosystem services varied according to the strategy decisions of forest owners, such as the type and level of implemented forestry operations. Institutional arrangements can constrain or facilitate particular strategies of forest management that supply different types and levels of ecosystem services [16]. As information on the potential impacts of economic policy measures on forest management is necessary for policy makers to design policy instruments, research on the preference of forestry operations and the appropriate level of compensation for forest owners participating in nature-friendly forestry is important. The results of such research can be utilized to design policy instruments that facilitate forest owners’ voluntary participation in activities enhancing ecosystem services that benefit the public. With the voluntary participation of private forest owners in nature-friendly forestry, the forestry sector can help meet the national targets for biodiversity conservation and climate change mitigation and adaption, as well as the societal demand for ecosystem services. In particular, there is a need for a policy design that can encourage private forest owners’ participation in sustainable forestry by incorporating the preferences of owners of forestry operations, subject to compensation for opportunity costs.

4.1. Restoration of Planted Forest to Natural Forest

Our results suggest that forest owners are willing to restore their forests to natural forests by half, whereas they are reluctant to transform all of their forests into natural ones. They demand KRW 185,280/ha/year for such operations. They might consider that the regeneration of preferred tree species may become more difficult after conservation in natural forests. Timber production can become less productive, and there is an opportunity cost for the conservation of natural forests. For example, maple syrup productivity in artificial forests is higher than that in natural forests in the ROK [48]. Similarly, Danish forest owners’ utility becomes negative when forests are left untouched, and thus require compensation [16]. In a previous study, forest owners were willing to allow five mature trees to decay naturally per hectare, even with negative WTA, similar to the option of choosing 50% restoration of natural forests in this study. According to Gonzalo et al. (2021), forest owners were asked to indicate their preferences among three levels of restoration (33, 66, or 90% restoration of their lands) regarding whether they would restore or maintain native vegetation cover. Their results revealed that they were more likely to prefer smaller proportions of biodiversity conservation using a mixed logit model [49].

To implement the payment for ecosystem services to compensate forest owners’ opportunity costs, the budget allocated for the current subsidy scheme could be reallocated instead. Currently, the South Korean government subsidizes forest owners for reforestation from 90 to 100% of the reforestation cost [50]. According to this study, private forest owners demand KRW 185,280/ha/year for a 100% conversion of their man-made forests to natural forests. This amount is more or less equal to the amount of plantation subsidy of KRW 9,323,000/ha, including labor and transportation costs, assuming that 3000 seedlings of Japanese larch were planted per hectare (“Creation and Management Of Forest Resources Act”). With a rotation of 50 years, the total WTA for rotation was KRW 9,264,000/ha. If the government budget for the reforestation subsidy is reallocated, it could cover the budget for compensation required for forest owners to restore artificial forests to natural ones.

For restoration to 50% natural forest activity, forest owners showed their utility positively. We assumed that this is because some of their forests have already been heading toward natural forests without their active efforts. The reasons are that some of the forest owners may prefer the natural forests since plantations are less profitable than natural forests [51,52,53,54], and more recently, as agroforestry producing wild vegetables, mushrooms, etc., in the natural forest has been popular, there are some forest owners considering restoration to natural forests that do not incur an opportunity cost. In addition, considering that absent forest owners who do not live in the same municipality of their forest holdings in Korea is 56% [24], some owners are already experiencing that their poorly managed man-made forests have been already restored to natural forests. Such absent or less interested forest owners do not seem to demand compensations for restoration to natural forests.

4.2. Extension of the Rotation Age of Timber Harvesting

We find that owners have a statistically negative utility for extending the rotation age of their forests by 50 years and require KRW 212,060/ha/year. However, they were not open to extending the rotation age by 20 years. This implies that forest owners are aware of the increasing value of timber quality through rotation extension. However, extending the harvesting age to 50 years can cost more than the increase in the timber value. Currently, most timber is harvested at the lower limit of the rotation age set by the KFS, and the harvested timber is used as wood chips for biomass energy, wherein the price of wood logs is approximately KRW 80,000/m³ (Korea Forestry Promotion Institute, 2020, https://www.kofpi.or.kr/eng/about/about_01.do, accessed on 8 January 2023).

Kline et al. (2000) examined the forest owners’ willingness to forego all timber harvest for 10 years for wildlife habitation in the states of Oregon and Washington in the United States (US) [13]. According to their study, forest owners demanded annual tax reductions ranging from approximately USD 62 to USD 2471 per hectare. Similar to our study, Matta et al. (2009) surveyed the willingness of forest owners in Florida, US, to restrict the harvesting age and only harvest after the trees were 50 years old to promote forest health and wildlife habitat, in contrast to the commercial harvest rotation age of 30 years; this had a significant effect on the decision making of forest owners [20]. The WTA was estimated to be USD 52.61/ha/year for this practice. Nunez Godoy et al. (2022) revealed that forest owners in Argentina did not prefer to enroll in conservation programs when restricted to sustainable timber harvesting and required a compensation up to USD 50/ha/year [22]. Annual payments requested by forest owners in each region may differ, and we found that they would not be satisfied with timber harvesting restrictions. In the Korean case, Park and Youn (2021) revealed the willingness of forest owners to accept when they extend the rotation age to 60 years and 100 years for carbon credits. It showed that forest owners demanded KRW 17,039/tCO₂ and KRW 23,070/tCO₂ [55]. Given that forest per ha mitigates 10.4 tCO₂ for a year based on 30-year-old trees [56], the results of WTA per ha and per year in both studies seemed similar.

Referring to the estimation of the annual standard carbon sequestration of major forest species per hectare [56] and carbon pricing, which was estimated as KRW 18,000/tCO₂ as of May 2021 through the Korea Emission Trading System, we can approximately estimate the upper limit of potential carbon sequestration. For example, if it is extended by 50 years from 25 years, which is the minimum rotation age set by the KFS regulation for sawtooth oak, the compensation for the opportunity cost of extending the harvesting age by 25 years could be substituted by a price of carbon credits of KRW 6.42 million per ha As the methodology of the extension of rotation age has not been accepted for ETS in ROK yet, here, we assumed that carbon credits certified for the practice of rotation extension could be traded and the price could be used as a surrogate of the opportunity cost.

Furthermore, there is an issue in international society related to the optimum rotation age considering the market and non-market values, such as carbon sequestration, watershed protection, and habitat for endangered species. However, in this study, we suggest that forest owners extend their rotation age as one of the forestry operations for public ecosystem services. Several respondents rejected the answer, as they believed that extending the rotation age incurs a considerably higher opportunity cost than the value of environmental services. Further investigation is needed on the value of extending the rotation age to accurately estimate the differences between market and non-market values.

4.3. Forest Protection Activity and Thinning

We hypothesized that forest protection activities not only enhance environmental services but also protect forest resources from thefts, wildfires, landslides, diseases, and insects. Therefore, we expected that forest protection activities could have a positive impact on forest owners’ utility, and they might not demand compensation for such activities. In contrast, residents living near forests in West Kalimantan and South Sumatera view the forest monitoring activity as inducing an opportunity cost of time as it requires labor, and they are occupied by their farms, planting coffee, and shifting cultivation. Thus, forest residents who practice agroforestry tend to ask for compensation when required to regularly protect forests [57].

In the case of thinning, for non-industrial forest owners in the southern US who practice commercial thinning to prevent southern pine beetles, thinning operations had no significant effect on their utilities [58]. In our study, Korean private forest owners did not demand compensation for thinning activity and were willing to practice it. This is because, in South Korea, there are government subsidies for thinning operations, which cover up to 100% of the thinning costs. Thus, we assumed that forest owners are willing to practice thinning as it could increase forest productivity without additional cost from their own pockets.

Although private forest owners in the ROK are positive about forest protection activities, they are not actively implementing forest protection. This is likely due to the high proportion of absent owners and the small size of forest holdings. The average size of forest holdings owned by private owners, accounting for 66% of total South Korean forests, is 1.9 ha per individual forest owner, and 85.7% of them are less than 3 ha [24]. As absent forest owners account for 56% of private forest lands, they face physical difficulties in managing their forests regularly. However, they can protect their assets through forest protection activities such as forest fire prevention and pest management if there is support from the government. Likewise, with thinning, even though it is difficult to frequently visit and manage forests, they seem to consider it is worthwhile to implement for the economic, social, and environmental values and benefits of forests if government subsidies are provided.

4.4. Policy Implications

The direct payment scheme for forestry business implemented by the KFS aims to support forestry farms whose household income is at the lowest level among all sectors, while providing environmental services. However, the scheme determines the payment amount based on the level of forestry operations engaged in forests, providing services such as timber and non-timber forest production.

The results show that private forest owners do not demand compensation for some forestry practices even though they can enhance nature’s contribution to people. These include the extension of the rotation age, protection activity once or twice a month, restoring half of their forest lots to natural forest, and thinning. In contrast, activities such as conversion to natural forests and the extension of the rotation age by 50 years must be compensated for their opportunity cost. The amount of compensation required, estimated in terms of the total MWTA, was KRW 397,340/ha/year. In South Korea, the government purchases private forests to increase public services of forests, including conservation of biodiversity, watershed protection, prevention of disasters such as floods and landslides, and promotion of forest welfare services such as recreational opportunities in forests. The KFS Eastern District office purchased private forests at a price of KRW 4,392,220/ha in 2022. Thus, we hypothesized that the interest in this amount of land price could be the upper limit of payment for ecosystem services to forest owners for their contribution to public benefits to encourage them to participate in environmentally friendly forestry. Although there are forest policies for forest owners and ecosystem enhancement services such as purchasing the private forestland and substitute management, PES should be used to encourage voluntary participation by forest owners. For example, if a forest owner creates additional public forest ecosystem services in addition to the production of timber and non-timber forest production, the government can compensate the foregone opportunity by calculating the additional service value and opportunity costs associated. At this time, the necessary budget can be brought from the budget of subsidies, which have been already executed for other policies supporting commercial forestry in private lands. We recommend that the government considers setting a standard of the PES scheme based on the WTA of private forest owners in meeting the demand for ES by the public. Accordingly, forest owners would be able to manage their forests such that their forestry business can yield a maximum of social surplus as well as maximizing their profits. For instance, he or she can extend the rotation age by ten years beyond the optimum rotation age, resulting in maintaining increased forest value in terms of not only financial returns but also ecosystem services, namely increased carbon stock in forest lands with compensations from the government. There have been a number of studies on the willingness to participate in the forest program and the factors affecting their will in the ROK [59,60,61,62,63]. However, there are few studies on the compensation amount that forest owners require for their participation in forestry for public benefits. According to a study by Koo et al. (2014), the government-led forest policy has a positive effect on private forest owners’ interests in PES scheme. Their study result indicates that the government should provide a clear guideline for forestry geared to public benefits and monitor forest owners’ opinions on the issue regularly [59].

4.5. Research Limitations

This study revealed the preferences of forest owners’ investment in forestry activities geared to the provision of public ecosystem services. As the subject of current research is limited to the members of Korea Forestry Successor Association, who are more active in forestry business compared to the normal forest owners, the findings may not represent the whole population of private forest owners. Therefore, further research is necessary to investigate the willingness of forest owners to invest in forestry business-enhancing public ecosystem services, not only covering forestry sessors that are officially recognized but also other types of forest owners.

5. Conclusions

Forests in Korea account for 63% of the total land area, of which private forests account for 67%. As individuals who own private forests account for 92% of all private forest owners, private forests have great potential to provide ecosystem services and achieve the national target for greenhouse gas emissions. Although the KFS implemented the DPSFB, some limitations still exist, as it does not consider the level of forestry operations geared toward public environmental services. This study was intended to contribute to designing payment schemes for ecosystem services for nature-friendly forestry by identifying forest operations that forest owners are willing to implement and the amount of compensation they are willing to accept. The main research findings are as follows. First, forest owners are willing to consider positive ecological restoration of up to 50% of planted forests to natural forests without the loss of utility, but their utility will be negatively affected in the case of restoring 100% of their land to natural forests. In this case, they require KRW 185,280/ha/year as compensation for the opportunity cost. We speculated that this is because the regeneration of preferred tree species would be difficult once their lands are converted to natural forests; thus, timber production would become less productive. Extending the rotation age by 50 years is negatively perceived at a compensation of KRW 212,060/ha/year, whereas extending the rotation age by up to 20 years beyond the minimum age stipulated by the regulation does not impact their profits in timber production. Additional years of waiting for the harvest beyond 20 years could incur financial costs for forest owners. Forest protection activities twice a month can yield some benefits to forest owners, such as preventing the loss of timber production and thinning operations that improve timber growth and quality. As forest protection and thinning operations involve both costs and benefits, forest owners can voluntarily implement these operations without compensation. Compensation for opportunity costs incurred due to forestry operations for public benefits, namely, 100% restoration to natural forests and the extension of rotation age by 50 years, could be obtained from the KFS and local governments’ budgets if forest policy is redesigned for PES schemes for ecosystem services instead of providing subsidies for forestry inputs without accounting for the public benefits specific to the type of forestry operations. The level of compensation required by private forest owners in the ROK is within the current level of the government budget allocated for reforestation subsidies or the purchase of private forests for enlarging national forests.

As voluntary participation is required, it is also necessary to balance the outputs of ecosystem services and the utility of forest owners arising from their actions. For example, prioritized payment for nature-friendly forestry operations could be offered to extend the rotation age by 50 years, as it could not only promote the ecosystem services, including biodiversity and climate mitigation, but also private interests of harvesting high-value timber at the end. The results of this study provide information on forest owners’ preferences for forestry operations regarding the enhancement of ecosystem services and their willingness to engage in environmentally friendly forestry. Moreover, they also highlight the implicit opportunity cost of providing additional ecosystem services for the public that can further be supplied with appropriate intervention by the government. Such information can be used to design payments for ecosystem services and direct payment schemes for private forestry in the ROK. If the current budget is relocated to PES schemes for nature-friendly forestry, it could change the strategies of forest management used by private forest owners, and subsequently the type and number of environmental services provided to the public. Further study is required on the preferences of forest owners to invest in forestry business-enhancing public ecosystem services, not only covering forestry sessors that are officially recognized, but also other types of forest owners.

Research on their preferences of compensation schemes include subsidy, direct payment for ES, and tradable permits to promote private forest management for public forest ecosystem services provision. Further research on the factors affecting the participation of forest owners in PES programs is necessary to provide information for policy makers in forest sectors.