Developing a microsimulation model for farm forestry planting decisions

There is increasing pressure in Europe to convert land from agriculture to forestry which would enable the sequestration of additional carbon, thereby mitigating agricultural greenhouse gas production.¹ The recognition of the multiple values of forests has led to international efforts to increase the planting of new forests (afforestation). However, despite the provision of economic incentives, afforestation targets across Europe are not being met (European Commission, 2013).² This policy failure has consequences downstream in the value chain in for timber processing, for increasing demand for wood fibre and for the potential of forests to sequester carbon and mitigate greenhouse gas emissions. In this paper, we develop a microsimulation model to help to understand the drivers of forest planting behaviour

This paper develops a microsimulation based framework looking at the drivers of decisions in relation to afforestation (new planting). The literature suggests that a number of factors affect the afforestation decision. In a meta-analysis (Beach et al., 2005) assessed the primary factors driving decision-making among landowners considering afforestation: owner characteristics, plot/resource conditions, factors that affect the forest investment decision and market drivers (costs and returns from forestry and alternative enterprises). Edwards and Guyer (1992) reported that the principal constraints to planting in the UK include lack of land, duration of the commitment and the inability of the annual subsidy payments to compete with agricultural returns. Van Gossum et al. (2012) and Moons and Rousseau (2007) suggest that Flemish forest subsidies are not high enough and the absence of financial benefits for farmers was one of the main reasons for the limited uptake by farmers in the Netherlands (Van Gossum et al., 2010).

With the growing importance of natural resource and environmental issues in policy spheres, the microsimulation literature has been developing in natural resource spheres such as agriculture, environment and land use (Richardson et al., 2014; Hynes and O’Donoghue, 2014; O’Donoghue, 2017; Ryan et al., 2014). In developing a microsimulation model with the aim of simulating the drivers of afforestation behavior on farms, this paper charts the development of one of the first forestry based microsimulation models.

While our interest in this study is on drivers of behaviour, we do not develop a behavioural simulation model, but one that is static in nature, where we develop a static microsimulation framework which takes a population distribution and simulates ‘day-after policy’ changes on farms. Thus we develop a model initially with a focus on drivers of forestry planting behaviour. It is therefore a forestry parallel to the microsimulation literature that looks at tax and social policy. In understanding pressures on behavior, this analysis develops static-based measures akin to the use of replacement rates (O’Donoghue, 2011) and marginal effective tax rates (Immervoll, 2005). In doing this, we model the impact on income and public transfers and other issues such as land value and hours worked as a result of converting agricultural land to forestry. We consider in particular the differential implications of planting across the distribution of farms, noting significant heterogeneity amongst farms. Further light could also be shed on the levers driving the land use change by specifically comparing the farm incomes and incentives for those farms that have planted in the past against the farm incomes and incentives that applied to farms without forests.

We select Ireland as a case study as it is the country within Europe that saw the most rapid increase in afforestation following the introduction of financial incentives for the afforestation of agricultural land. The period 1985 to 1995 saw a rapid increase in annual planting by farmers from just over 5,000 ha to almost 24,000 ha. In recent years however, the decline in afforestation rates has equally been most pronounced in Ireland relative to other European countries as the planting rate has fallen off to just under 5,000 ha in recent years(Forest Service, 2018), despite substantial increases in subsidy payments. Studies that specifically examine farmer afforestation in Ireland have found that there is a preference not to plant land that is profitable in agriculture (Frawley and Leavy, 2001; O’Leary et al., 2000; McDonagh et al., 2010; Ní Dhubháin and Gardiner, 1994; Duesberg et al., 2013; Duesberg et al., 2014) and a strong preference for the utility or value that farmers derive from the farming lifestyle. Increasingly, the loss of flexibility of land use as a result of the permanence of forestry is seen as a barrier to afforestation (McDonagh et al., 2010; Duesberg et al., 2014; Duesberg et al., 2013).

This paper develops a model to simulate financial and lifestyle drivers of the forestry planting decision. The analysis utilises farm micro data together with bio-economic forestry models and forest subsidy models to estimate the return from an alternative agricultural land use. An examination of the combination of these data and factors has not previously been undertaken in the forestry literature and is facilitated here by the use of microsimulation modelling, which abstracts from reality to better understand complexity (O’Donoghue, 2014). In Section 2 the factors influencing a long-term participation decision such as forestry are investigated in order to develop a theoretical framework. Section 3 describes the generation of the variables used in the models and Section 4 presents the results. We finish with discussion and conclusions.

In determining the nature of the microsimulation model required to simulate the drivers of afforestation behaviour, we firstly examine the theoretical framework required to compare the incentives of alternative land uses. The literature that addresses the land use change from agriculture to forestry (or farm afforestation) is actually quite limited as the vast majority of studies in relation to the economics of forestry focus on deforestation or management decisions in pre-existing forests.

As the return from forestry is quite long term with potentially a 40–50 year gap between planting and harvesting for coniferous trees (and longer (up to 150 years) for deciduous trees), afforestation is a long-term investment. Even with the long term planning horizon, afforestation is an unusual land use change as it is a permanent change in many countries (there is a legal requirement to re-plant the land after timber harvesting (Ryan et al., 2016). Conversely, the alternative land use, agriculture, has a relatively short time horizon with a one to three year cycle for agricultural produce. Thus the land use change incurs an inter-temporal annual opportunity cost foregone from existing farm enterprises such as dairy, cattle rearing, sheep or tillage enterprises. In comparing alternative land uses over a long period of time, we will need to model alternative income streams over a long time period and the consequential differences in the net present value of income.

The literature that deals with agricultural opportunity costs in the context of farm afforestation is also limited and focuses largely on the calculation of the opportunity cost of carbon sequestration through afforestation at an aggregate level (see Moulton and Richards, 1990; O’Leary et al., 2000). Due to the policy complexity of comparing agricultural and forestry income streams over time, studies in this area are based largely on average opportunity costs across farm systems. O’Connor and Kearney (1993), McCarthy et al. (2003) and Ryan et al. (2014), use average values to highlight the importance of the loss of agricultural subsidies when land is planted. In calculating the net return from farm afforestation, Dudek and LeBlanc (1990) use average agricultural opportunity costs and Breen et al. (2010) use average farm management data (Teagasc, various years) to calculate the opportunity cost of the superseded agricultural enterprise, while Upton et al. (2013) compare average life-cycle agricultural and forest incomes. Herbohn et al. (2009) go further and use actual farm input data to calculate the agricultural opportunity cost of planting on individual farms. However, in reality there is a large degree of income heterogeneity at individual farm level, yet to our knowledge, there are no studies that examine the distribution of farms in quantifying the life-cycle financial drivers that influence planting behaviour.

As farm afforestation is essentially a permanent land-use change which imposes loss of flexibility in relation to land use (as there is a legal obligation to retain land under forest once planted), we expect that planting could negatively impact on the long-term wealth of farmers, as planted land is no longer available for agriculture for current or future generations. Therefore in addition to changes in the net present value of income, we need also to consider changes in farm wealth, by examining whether there is a difference in farmers’ perception of the value of their land before and after planting.

Traditional economic theory suggests that individuals make decisions based on the expected change in their level of ‘well-being’, where the term used for well-being or welfare is utility (Edwards-Jones, 2006). Given that utility is a difficult concept to measure, economists have often made the simplifying assumption that money can act as a substitute for utility. This has led to the situation observed in many agricultural economic models in which farmers act to maximise profitability in all circumstances (Edwards-Jones, 2006). However, the Random Utility Maximisation (RUM) and Discrete Choice methodologies employed by McFadden, 1973 and Ben-Akiva and Lerman (1985) incorporate the value of leisure time as a component of utility.

Labour requirements also differ between agricultural and forest land use. Relatively little labour is required for forest management on a day to day basis, whereas for an animal based agricultural system, there are regular day-to-day labour requirements, e.g. feeding and maintaining the animals and grassland management. Thus we expect that farmers who plant land have reduced working hours with a consequent increase in potential leisure time period. This may act as an incentive for farmers who are motivated by life-style and may be willing to make a trade-off between income and leisure (Ryan et al., 2015).

Previous studies have highlighted the importance of the relative potential incomes from agriculture and forestry. However, the impact of planting on annual income is less clear because the sources of income are so different. Agriculture and forestry are highly regulated and subsidised businesses, given the public good nature of many of the outputs. Income is thus derived from both market and public sources. Agricultural income includes annual market income and agricultural policy subsidies, which are tied either to production in the case of coupled payments, or land in the case of decoupled payments.³ As planting land results in a reduction in the agricultural land area, agricultural subsidy payments may be reduced on planting. The opportunity costs of planting arising from the loss of agricultural market and subsidy income are intertemporal in nature as they are essentially incurred for each year of the forest rotation.

Income from forestry is net of up-front planting costs, regular maintenance and management costs,with periodic thinning costs. Income may arise from periodic thinnings with harvesting income arising at the end of the forest rotation. Afforestation income streams also include considerable policy subsidies. Thus, losing land from agriculture to forestry, in a straight land use change, sees a reduction in both agricultural market incomes and subsidies and ‘lumpy increases in forestry market incomes and subsidies. The time-lines of the market and subsidy income, along with the wealth and leisure components of the farm afforestation decision are important in developing the inter-temporal framework and are presented graphically in Figure 1.

Figure 1

Download asset Open asset

Apart from the high-level income, wealth and leisure differences between agriculture and forestry, there is also significant micro-unit variability as farms can vary by soil type and agronomic characteristics (which influence productivity), farm system (which influences incomes and costs) and individual farmer characteristics and skills (which influence efficiency and investment decisions). Thus we expect the drivers of the planting decision to vary considerably across farm types.

In summary, the inter-temporal modelling of the differential life-cycles of the agricultural and forest income cycle factors addressed in the literature should provide a better understanding of the direct and indirect opportunity costs faced by individual farmers in considering forestry. However, due to the dearth of farm forestry micro-data, previous studies have been undertaken either using aggregate data or sectoral averages, yet these do not consider the differential farm and farmer characteristics and environmental conditions that apply at individual farm level. Thus an analytical, microsimulation based approach that provides information on the economic drivers of planting behaviour - across the distribution of farms - is appropriate to model alternative land use decisions and assess the impact of financial and lifestyle drivers.

This study develops a model that simulates at a micro-level, the market, environmental, policy and leisure drivers of forestry planting, in an inter-temporal framework. Here we describe the methodology used to develop a farm-forest microsimulation model which allows us to compare income streams over time between the alternative land uses of agriculture and forestry.

The output from the static microsimulation model initially provides forest market and subsidy income simulations derived using ForBES, along with land value and hours worked values derived from NFS farm micro-data.

ForBES provides cost and income curves for thin and no-thin scenarios and generate annual equivalent NPVs for a range of discount rate, subsidy, thinning, rotation and indexation options. The life-cycle pattern of costs and incomes are presented graphically in Figure 2 Figure 2for establishing a Sitka spruce (SS) forest (which accounts for almost two thirds of non-industrial private forestry in Ireland) in 2015, across a range of productivity (yield) classes (yc).

Figure 2

Download asset Open asset

There is a considerable difference between the curves for the conifer thin (ct) and no-thin (cnt) scenarios. The distribution of income and cost in the early years of the rotation is the same for both thinned and unthinned crops, regardless of yield class, as early income is derived from grant and premium subsidies which are categorised only by species and not by site productivity. Across the yield classes, there is an almost incremental increase in income.

In this analysis, the reforestation cost of the next rotation is included as a cost in the first rotation and is evident in the substantial increase in costs after clearfell. While it can be argued that theoretically, the cost of reforestation should not be apportioned to the first rotation, from a pragmatic and transparency perspective, forest owners cannot achieve clearfell revenues without incurring the cost of reforestation. In reality, because the cost arises at the end of the rotation and is dwarfed by the clearfell revenue, it is unlikely to have much impact on the magnitude of the NPV.

The purpose of this analysis is to try to understand the impact of planting on the relativity of agriculture and forestry incomes and the impact on land values and hours worked. In our data, we observe the farms that have planted (has forest) and those that have not (no forest). In order to understand the drivers of forest planting behaviour, we first simulate the impacts of a marginal change in land use. We utilise the Teagasc ForBES and the ForSubs models to simulate forest market and subsidy income respectively and utilise the two reduced form econometric models described above to simulate counter factual land and labour impacts. However these averages mask significant heterogeneity across farms so we also disaggregate farms on the basis of whether the agricultural or forest income life cycle streams are greater over time for each farm in the population. Variables are generated whereby life cycle forest income streams are defined as annual equivalised (AE) NPV of market plus subsidy income and compared directly with agricultural incomes (on a per hectare basis). Agricultural life cycle income streams are defined as AE (NPV) of market income, farm gross margin and farm net margin (on a per hectare basis). These impacts are examined for both ‘has forest’ and ‘no forest’ farms. This categorisation is at the heart of this analysis as it is expected that the relativity of these life cycle incomes is a major driver in the afforestation decision.

Table 1 describes the average changes in income, land and labour which would result from the decision to plant a proportion of the farmland, simulating in turn, a straight land use change of five percent of the land, the change in value and the change in hours worked.

We see that for each of the income types (market, gross margin and net margin), planting five percent of the farmland results in a slight reduction in income, (as expressed in terms of AE of NPV), as the income from forestry is lower on average than agricultural income. Hours worked reduces due to the lower time input required for forestry than for livestock farming, thus increasing leisure. Although the reduction in hours is small, this is consistent with the underemployment in Irish agriculture noted by Loughrey and Hennessy (2016). Land values show a small decrease on average due to preferences for farming and lower land flexibility, however given the contribution of land to the overall asset base, a 3%–5% fall in value is quite important.

To test the sensitivity of the results, we utilise data from 2012 to 2015 to compare the ratios for farms with forests against those without forests. The results are presented in Table 2 where farms with higher incomes from forestry than from agriculture (For> Ag) are also examined on the basis of the three agricultural income definitions. Finally, future planting intentions are examined, comparing farms that indicated they would ‘never plant’ (1) with those who might plant (never plant 0).

The proportion of farms that have a higher AE for forestry than agriculture is greater amongst those who have planted forests. This is true for each measure of income considered. Considering market income, 63% of farmers who planted had a higher AE for forestry than for agriculture. However the rate for those who have not planted forestry is 53%. Comparing forestry returns with the agricultural gross margin, where agricultural subsidies are included, the proportion falls to below 50%, with rates of 30% and 45% respectively for farms with and without forestry. When subtracting overhead costs (net margin), the share peaks at 70% and 60% respectively.

Turning to future intentions, we find similar conclusions. Of those who ‘might plant (corollary of never plant) , there are more farmers who have a higher AE from forestry than from agriculture. However as is the case for those farms that have already planted, the gap is relatively low, again indicating the heterogeneity of the underlying behavior.

Overall, there are substantial numbers of farmers who would be financially better off by planting, but who have not done so, while similarly there are substantial numbers of farmers who have planted, even when they would have been better off to retain all their land in agriculture. The former may be explained as a result of cultural factors underlying a reluctance to plant (Ní Dhubháin and Gardiner, 1994; Frawley and Leavy, 2001; Malone, 2008; O’Leary et al., 2000; McDonagh et al., 2010; Duesberg et al., 2013), while the latter cohort may be influenced by non-pecuniary factors.

In Table 3, we decompose the annual equivalised NPV into income components (average) for farms that have planted and those that have not. Where agriculture has a higher AE than forestry (Group A), the biggest difference is evident in farm income (output - direct costs), which is substantially different to those farms with a lower agricultural AE (Group B). This is partly due to the substantially higher incomes on dairy versus other systems and the fact that dairy farms are proportionally located on productive soils. Farm subsidies are higher for Group A but the gap is quite small. Overhead costs, consistent with productivity, have a similar pattern to farm incomes. The gap for forest income is relatively small, with potential forest income being slightly higher when the agricultural income is higher, as both agricultural and forest incomes are greater on more productive soils.

When we compare ‘has forest’ farms in Group A with ‘no forest’ farms, all income differences (forest – agriculture) are negative, but less negative for ‘has forest’ farms. The biggest driver of difference is lower agricultural incomes. For group B, the forest-agriculture income difference is higher for ‘has forest’ farms. Again the differential is primarily driven by farm income. In fact potential forest market income and subsidies are higher for ‘no forest’ farms, suggesting that forest policy is relatively untargeted from an incentive perspective.

In Table 4 we take the difference between forest and agricultural market incomes and group this difference into deciles, where higher deciles have a higher difference between forest and agricultural income. Plotting the probability of having a forest, we find that for ‘has forest’ farms, there is a substantial difference between the biggest gap (32%) and the lowest decile with only 7% difference. In general however, we see a relatively small difference in the planting rate between the 2nd and the 8th decile, indicating that financial incentives seem to only have an impact at the extremes of the distribution.

The probability of never planting takes as expected, the opposite trend, being greater for the lowest gap and least for the highest gap. However the range between the top and the bottom is not as great proportionally as for ‘has forest’ farms.

While Table 4 considered probabilities for market incomes only, we next expand the analysis to examine the deciles across all three agricultural income measures for ‘has forest’ farms. The results presented in Table 5 indicate that while the trend is similar, the proportional range between decile 1 and 10 is smaller for the gross margin and net margin income measures. Thus, it would appear that the planting is more strongly influenced by the financial gain associated with the market income measure.

In Table 6, we examine the deciles of the difference between forest and agricultural incomes (forest minus agriculture) in relation to the individual components of income. The difference in forest income between decile 1 and 10 is relatively small for farm subsidies, with the biggest difference being due primarily to agricultural market incomes and to a lesser extent, to overhead costs. Thus it appears that farm income variability has a greater impact on financial drivers than forestry incomes. This raises an important policy question as to whether forest subsidies should be paid on a straight per hectare basis, or on the basis of the opportunity cost foregone from the superseded land use.

Finally, Table 7 describes the characteristics of farms by decile. The lowest deciles with the largest agriculture-forestry gap are more likely, unsurprisingly to have a greater share of higher income. These farmers are likely to be younger dairy farmers with more intensive systems and higher stocking rates, are more likely to pay for extension services, to be full time farmers and farming on better soils. The corollary also holds that those with a greater financial incentive to plant forests are more likely to be older, farming on marginal land in drystock systems and are more likely to work off-farm.

This paper describes a static microsimulation approach which provides a better understanding of the life-cycle relativity of forestry and agricultural incomes, using Ireland as a case-study. Due to the complexity and micro data requirements, there is a general gap in the land use change literature in relation to modelling at the individual farm level. The unique contribution of the microsimulation approach taken in this paper is that it combines two units of micro analysis, namely the farm and the forest. While there are existing studies in the literature that individually model farm and forest level incomes, studies that examine the incomes from both land-uses are limited. However, information on both land uses is necessary to understand such a major and permanent land use change. This analysis examines both private (market) and public (subsidy) income sources arising from farming and forestry activities. In combination, these different income sources produce a high degree of heterogeneity, leading to very detailed farm level results.

The static microsimulation methodology allows for the generation of actual and counterfactual forest and agricultural income streams and for other attributes of utility such as long-term wealth and leisure, for the first time. Income variables are generated to directly compare the life-cycle income streams from agriculture and forestry (on a per hectare basis) given the agronomic and environmental conditions of individual farms.

The results show that financial drivers are significant and important drivers of farm afforestation as the proportion of farms with greater life-cycle income streams from forestry is higher amongst farmers who have already planted. The results also show that substantial numbers of farmers would be financially better off to plant forests, however many of these farmers have not planted. Malone (2008) recognizes that the decision to change from agriculture to forestry on a parcel of land is not taken in isolation and lists multiple factors including personal circumstances, the relative attraction of available subsidies and the permanent nature of the decision, which removes other options for land use and has implications for current and future generations. The static microsimulation methodology utilised in this study to quantitatively determine the economic implications of planting referred to by Malone (ibid.), has not previously been used in the agriculture to forestry land use change context.

While previous afforestation incentives focused only on horizontal policies, this analysis suggests that policies that are targeted to take account of direct (financial) and non-direct (lifestyle and land use flexibility) drivers of afforestation, may be more effective in increasing the uptake of farm afforestation. It is evident that the differential between potential forest and agricultural income is driven primarily by farm income, while planting behaviour is more strongly influenced by the financial gain associated with the market income measure. This raises an important policy question as to whether forest subsidies should be paid on a straight per hectare basis, or on the basis of the opportunity cost of the alternative land use. It would further appear that forest policy is relatively untargeted from an incentive perspective and that the financial incentives seem to only have a strong impact at the extremes of the distribution of farms in relation to income potential. This suggests that targeted incentives might be more effective than the horizontal incentives that have previously applied across all farms irrespective of system and income levels.

The average decrease in land value reported by farmers after planting reflects the inter-temporal commitment and loss of flexibility of land use associated with afforestation. While the permanent nature of forestry as a land use has been cited as a barrier to planting (McDonagh et al., 2010) the decrease in self-assessed land value has not previously been quantified.

As the afforestation of agricultural land is now a primary component of the climate change mitigation tool kit, methodologies that provide greater understanding of individual farm level economic drivers that could be manipulated to incentive the necessary land use change are increasingly important. The methodology presented here is relevant for Ireland, but the framework employed uses Irish FADN data collected by the Teagasc National Farm Survey and thus has potential for cross country comparative analysis using FADN data. Thus the methodology is readily scaleable to the EU level and potentially globally. In the broader context of Climate Smart Agriculture and the Grand Challenges facing the intensification of agricultural production, these findings have implications for policies that seek to optimise natural resource use.

1. 1
   Econometric studies of non-industrial private forest management: a review and synthesis

   1. RH Beach
   2. SK Pattanayak
   3. J-C Yang
   4. BC Murray
   5. RC Abt

   (2005)

   Forest Policy and Economics 7:261–281.

   https://doi.org/10.1016/S1389-9341(03)00065-0

   • Google Scholar
2. 2
   Discrete Choice Analysis: Theory and Application to Travel Demand

   1. ME Ben-Akiva
   2. SR Lerman

   (1985)

   Cambridge, Ma: MIT Press.

   • Google Scholar
3. 3
   Irish land use change and the decision to afforest: an economic analysis

   1. J Breen
   2. D Clancy
   3. M Ryan
   4. M Wallace

   (2010)

   Irish Forest 67:6–20.

   • Google Scholar
4. 4
   Economics of Irish Forestry: Evaluating the Returns to Economy and Society

   1. JP Clinch

   (1999)

   Dublin: COFORD.

   • Google Scholar
5. 5
   Offsetting New CO2 Emissions: A Rational First Greenhouse Policy Step

   1. J Dudek
   2. A LeBlanc

   (1990)

   Contemporary Policy Issues 8:29–41.

   • Google Scholar
6. 6
   Assessing policy tools for encouraging farm afforestation in Ireland

   1. S Duesberg
   2. Áine Ní Dhubháin
   3. D O’Connor

   (2014)

   Land Use Policy 38:194–203.

   https://doi.org/10.1016/j.landusepol.2013.11.001

   • Google Scholar
7. 7
   To plant or not to plant—Irish farmers’ goals and values with regard to afforestation

   1. S Duesberg
   2. D O’Connor
   3. Áine Ní Dhubháin
   4. A Ní Dhubháin

   (2013)

   Land Use Policy 32:155–164.

   https://doi.org/10.1016/j.landusepol.2012.10.021

   • Google Scholar
8. 8
   International accounting standard IAS 41 (agriculture)

   1. EC

   (2009)
9. 9
   Modelling farmer decision-making: concepts, progress and challenges

   1. G Edwards-Jones

   (2006)

   Animal Science 82:783–790.

   https://doi.org/10.1017/ASC2006112

   • Google Scholar
10. 10
    Yield Models for Forest Management

    1. PN Edwards
    2. JM Christie

    (1981)

    Yield Models for Forest Management, Forest Commission Booklet 48, HMSO, London.

    • Google Scholar
11. 11
    Farm woodland policy: an assessment of the response to the farm woodland scheme in Northern Ireland

    1. C Edwards
    2. C Guyer

    (1992)

    Journal of Environmental Management 34:197–209.

    https://doi.org/10.1016/S0301-4797(05)80151-3

    • Google Scholar
12. 12
    Ministerial conferences

    1. European Commission

    (2013)
13. 13
    Berechnung des Werthes, welchen Waldboden, sowie noch nicht haubare Holzbestände für die Waldwirthschaft besitzen

    1. M Faustmann

    (1849)

    Allgemeine Forst- und Jagd-Zeitung 25:441–455.

    • Google Scholar
14. 14
    Afforestation Statistics

    1. Forest Service

    (2018)

    Department of Agriculture Food and Marine.

    • Google Scholar
15. 15
    Forest statistics

    1. Forestry Commission

    (2013)

    http://www.forest.gov.uk/website/forstats2013, [Accessed Jan 2014].

    • Google Scholar
16. 16
    Farm forestry: land availability, take-up rates and economics

    1. J Frawley
    2. A Leavy

    (2001)

    Teagasc: RERC.

    • Google Scholar
17. 17
    Social security money’s worth

    1. G Geanakoplos
    2. SO Mitchell
    3. SP Zeldes

    (1998)

    Social security money’s worth, NBER Working Paper Series 6722.

    • Google Scholar
18. 18
    The Australian farm forestry financial model

    1. J Herbohn
    2. N Emtage
    3. S Harrison
    4. D Thompson

    (2009)

    Australian Forestry 72:184–194.

    https://doi.org/10.1080/00049158.2009.10676300

    • Google Scholar
19. 19
    Woodland Management

    1. WE Hiley

    (1954)

    Faber & Faber.

    • Google Scholar
20. 20
    Handbook of Microsimulation Modelling

    1. S Hynes
    2. C O’Donoghue

    (2014)

    449–477, Environmental Models, Handbook of Microsimulation Modelling, Emerald Group Publishing Limited.

    • Google Scholar
21. 21
    Average and marginal effective tax rates facing workers in the EU: A micro-level analysis of levels, distributions and driving factors

    1. H Immervoll

    (2004)

    Average and marginal effective tax rates facing workers in the EU: A micro-level analysis of levels, distributions and driving factors, OECD Social, Employment and Migration Working Papers, No. 19.

    • Google Scholar
22. 22
    FALLING up the stairs: the effects of "bracket creep" on household incomes

    1. H Immervoll

    (2005)

    Review of Income and Wealth 51:37–62.

    https://doi.org/10.1111/j.1475-4991.2005.00144.x

    • Google Scholar
23. 23
    Annuity markets in comparative perspective: do consumers get their money’s worth?

    1. E James
    2. D Villas

    (2000)

    Annuity markets in comparative perspective: do consumers get their money’s worth?, Policy Research Working Paper 2493.

    • Google Scholar
24. 24
    Extending the boundaries of forest valuation

    1. S Kant

    (2003)

    Forest Policy and Economics 5:39–56.

    • Google Scholar
25. 25
    Handbook of Microsimulation Modelling

    1. J Li
    2. C O’Donoghue
    3. J Loughrey
    4. A Harding

    (2014)

    47–75, Static Models, Handbook of Microsimulation Modelling, Contributions to Economic Analysis, Volume 293, Emerald Group Publishing Limited.

    • Google Scholar
26. 26
    Farm income variability and off-farm employment in Ireland

    1. J Loughrey
    2. T Hennessy

    (2016)

    Agricultural Finance Review 76:378–401.

    https://doi.org/10.1108/AFR-10-2015-0043

    • Google Scholar
27. 27
    Factors affecting afforestation in Ireland in recent years. Report for the Minister of state with responsibility for forestry

    1. J Malone

    (2008)
28. 28
    Economic determinants of private afforestation in the Republic of Ireland

    1. S McCarthy
    2. A Matthews
    3. B Riordan

    (2003)

    Land Use Policy 20:51–59.

    https://doi.org/10.1016/S0264-8377(02)00052-2

    • Google Scholar
29. 29
    Missed opportunity or cautionary steps? farmers, forestry and rural development in Ireland

    1. J McDonagh
    2. M Farrell
    3. M Ryan
    4. M Mahon

    (2010)

    European Countryside 4:236–251.

    • Google Scholar
30. 30
    Conditional Logit analysis of qualitative choice behaviour

    1. D McFadden

    (1973)

    Frontiers of Econometrics.

    • Google Scholar
31. 31
    Policy options for afforestation in Flanders

    1. E Moons
    2. S Rousseau

    (2007)

    Ecological Economics 64:194–203.

    https://doi.org/10.1016/j.ecolecon.2007.02.021

    • Google Scholar
32. 32
    Costs of sequestering carbon through tree planting and forest management in the U.S

    1. R Moulton
    2. K Richards

    (1990)

    Washington DC: U.S. Department of Agriculture, ForestService, Gen. Tech. Rep. WO-58, December 1990.

    • Google Scholar
33. 33
    The participation decision in the rural environment protection scheme (REPS): a conditional logit approach using simulated counterfactual data for farmers

    1. G Murphy
    2. C O’Donoghue
    3. S Hynes
    4. E Murphy

    (2014)

    Paper presented to European Association of Agricultural Economists Congress.

    • Google Scholar
34. 34
    Farmers’ attitudes to forestry

    1. Á Ní Dhubháin
    2. J Gardiner

    (1994)

    Irish Forestry 5:21–26.

    • Google Scholar
35. 35
    Do tax-benefit systems cause high replacement rates? a decompositional analysis using EUROMOD

    1. C O’Donoghue

    (2011)

    Labour 25:126–151.

    https://doi.org/10.1111/j.1467-9914.2010.00501.x

    • Google Scholar
36. 36
    Farm-Level Microsimulation Modelling

    1. C O’Donoghue

    (2017)

    Farm-Level Microsimulation Modelling, Palgrave Macmillan, ISBN:, 978-3-319-63978-9.

    • Google Scholar
37. 37
    Afforestation in Ireland — regional differences in attitude

    1. TN O’Leary
    2. AG McCormack
    3. JP Clinch

    (2000)

    Land Use Policy 17:39–48.

    https://doi.org/10.1016/S0264-8377(99)00036-8

    • Google Scholar
38. 38
    Economic issues in Irish forestry

    1. R O’Connor
    2. B Kearney

    (1993)

    Journal of the Statistical and Social Inquiry Society of Ireland XXVI:179–209.

    • Google Scholar
39. 39
    The Handbook of Microsimulation Modelling. Contributions to Economic Analysis, Vol 293

    1. C O’Donoghue

    (2014)

    (Ed), The Handbook of Microsimulation Modelling. Contributions to Economic Analysis, Vol 293, ISBN:, 978-1-78350-589-2.

    • Google Scholar
40. 40
    A guide to the valuation of commercial forest plantations

    1. H Phillips
    2. D Little
    3. T McDonald
    4. J Phelan

    (2013)

    A guide to the valuation of commercial forest plantations, Dublin, COFORD, ISBN:, 978-1-902696-72-0.

    • Google Scholar
41. 41
    Tightropes and tripwires: new labour’s proposals and means-testing in old age

    1. Katherine Rake
    2. J Falkingham
    3. M Evans

    (1999)

    CASE paper 23, Centre for Analysis of Social Exclusion, London School of Economics.

    • Google Scholar
42. 42
    Handbook of microsimulation modelling

    1. JW Richardson
    2. T Hennessy
    3. C O’Donoghue

    (2014)

    505–534, Farm level models, Handbook of microsimulation modelling, Emerald Group Publishing Limited.

    • Google Scholar
43. 43
    The role of subsidy payments in the uptake of forestry by the typical cattle farmer in Ireland from 1984 to 2012

    1. M Ryan
    2. M McCormack
    3. C O’Donoghue
    4. V Upton

    (2014)

    Irish Forestry 71:92–112.

    • Google Scholar
44. 44
    Modelling financially optimal Afforestation and forest management scenarios using a Bio-Economic model

    1. M Ryan
    2. C O’Donoghue
    3. H Phillips

    (2016)

    Open Journal of Forestry 06:19–38.

    https://doi.org/10.4236/ojf.2016.61003

    • Google Scholar
45. 45
    Land use change from agriculture to forestry: a structural model of the income and leisure choices of farmers

    1. M Ryan
    2. C O’Donoghue
    3. V Upton

    (2015)

    International Association of Agricultural Economists. 2015 Conference.

    • Google Scholar
46. 46
    Management Data for Farm Planning. Annual Editions. (1984 to 2012)

    1. Teagasc

    Various years, Management Data for Farm Planning. Annual Editions. (1984 to 2012), Oak Park, Carlow, Teagasc.

    • Google Scholar
47. 47
    An analysis of money’s worth ratios in Chile

    1. C Thorburn
    2. R Rocha
    3. M Morales

    (2007)

    Journal of Pension Economics and Finance 6:287–312.

    https://doi.org/10.1017/S1474747207003150

    • Google Scholar
48. 48
    Examining the relative competitiveness of milk production: an Irish case study (1996–2004)

    1. FS Thorne
    2. W Fingleton

    (2006)

    Journal of International Farm Management 3:49–61.

    • Google Scholar
49. 49
    The potential economic returns of converting agricultural land to forestry: an analysis of system and soil effects from 1995 to 2009

    1. V Upton
    2. M Ryan
    3. N Farrelly
    4. C O’Donoghue

    (2013)

    Irish Forestry 70:61–74.

    • Google Scholar
50. 50
    Implementation of the forest expansion policy in the Netherlands in the period 1986–2007: decline in success?

    1. P Van Gossum
    2. B Arts
    3. J van Laar
    4. K Verheyen

    (2010)

    Land Use Policy 27:1171.

    https://doi.org/10.1016/j.landusepol.2010.03.007

    • Google Scholar
51. 51
    “Smart regulation”: Can policy instrument design solve forest policy aims of expansion and sustainability in Flanders and the Netherlands?

    1. P Van Gossum
    2. B Arts
    3. K Verheyen

    (2012)

    Forest Policy and Economics 16:23–34.

    https://doi.org/10.1016/j.forpol.2009.08.010

    • Google Scholar

Author details

1. Mary Ryan

   Teagasc, Rural Economy & Development Programme, Athenry, Co. Galway, Ireland

   For correspondence

   Mary.ryan@teagasc.ie

   Competing interests

   No competing interests reported

   "This ORCID iD identifies the author of this article:" 0000-0001-8395-6953

2. Cathal O’Donoghue

   NUIG Policy Lab, National University of Ireland, Galway, Ireland

   For correspondence

   Cathal.odonoghue@nuigalway.ie

   Competing interests

   No competing interests reported

   "This ORCID iD identifies the author of this article:" 0000-0003-3713-5366