The preceding three chapters have indicated circumstances under which tree establishment could significantly reduce flooding, sediment yield, or terrestrial erosion in New Zealand. These are summarised as follows:
Erodible lowlands and hills (but not mountains or catchment headwaters with naturally very high erosion rates).
Areas where there are frequent high intensity rainfalls (storms).
Land use / human settlement
Areas in grass or tussock cover prior to afforestation.
Areas of planted forest or scrub reversion or bush regeneration established more than
5–10 years (threshold date varies).
Areas of spaced tree planting on erodible hill country that is still farmed – applies only to reduction of terrestrial erosion effects.
Areas upstream of urban or rural infrastructure or intensive agriculture on lowland floodplains or other flood-prone landforms – applies only to reduction of sediment yield effects.
These are the characteristics of the terrain that could most usefully be targeted in afforestation and reversion policies.
This chapter reviews available information about the benefits and costs of these reductions. Where unavailable, it proceeds to discuss methods whereby the information might be obtained (Chapter 6).
Concepts for valuation
The basic approach to environmental valuation is to place values on the services that the natural environment provides. These services have value to society because people derive utility from their use, either directly or indirectly. People also value ecosystem services they are not currently using.
This approach to valuation is utilitarian (anthropocentric), in that any value is derived from the utility people derive from the services. Other values exist such as socio-cultural and intrinsic values; while these should be considered in resource management decisions, no tools to quantify these values have been developed.
The approach to the valuation of benefits resulting from policies or programmes follows the following steps:
- Identify the changes to ecosystem services brought about by decisions, policies, or programmes (eg, to plant forest that lead to changes in erosion, sedimentation and water yield).
- Quantify these changes using physical measures (areas saved from erosion, tonnes of sediments).
- Quantify the impacts caused by these changes on ambient environmental quality (increase flood risk, water quality) and environmental services.
- Determine the impact on capital, humans and animals (damage, recreation, aesthetics).
- Place a value on those impacts.
A typology for impacts14 (Step 4 above) include:
- increased flood severity
- water damage to farmland and settlements
- infrastructure damage (to bridges, roads, power supply)
- livestock losses
- decreased channel flood capacity
- silt damage to farmland and settlements
- increased flood severity
- reduced water quality (processing, recreation, fishing)
- biological degradation (habitat loss)
- water storage loss for irrigation and hydro-electric dams
- navigation (need for increased dredging etc)
soil erosion effects:
- loss of agricultural production
- on-farm damage to infrastructure (tracks, fences)
- increased conservation spending (private and public)
- direct property damage (residential)
- off-farm infrastructure (road/rail, bridges, utility network)
- recreational facility damage
- other effects, such a loss in farmer confidence, visual effects, etc.
The previous chapters have dealt with the first of these three steps. They have identified and quantified, where possible, the impact (in terms of benefits or costs) of afforestation on water yield, sediment yield, and soil erosion.
This chapter discusses valuation of those impacts (benefits and costs). The measure of value taken in this approach is peoples’ willingness to pay. For many goods and services, this value is reflected in what people are willing to pay in the market, and prices therefore reflect value. Many ecosystem services and impacts, however, are not traded in the market – even though they clearly affect society’s well-being ie, they reflect real costs and benefits.
To measure non-market goods and services, alternative approaches have been developed to discover the values people hold for them. Non-market valuation techniques can be divided in two main groups:
a. Revealed preference approaches. Information is obtained on people’s buying behaviour, or substitutes and complements of environmental goods and services. That information is used to derive a value for them. Approaches such as the Travel Cost Method, Hedonic Price and Defensive Expenditure fall in this category.
b. Stated preference approaches. People are directly asked (through surveys) about their willingness to pay or willingness to accept compensation for changes in environmental goods and services. Approaches such as Contingent Valuation and Choice Methods fall in this category.
Monetary valuation of the benefits / costs of climate change mitigation measures will require the use of a mixture of non-market valuation techniques, and cost approaches.
A fundamental difficulty with the estimations of monetary values requested in MfE’s brief for this review is the specificity of the estimates required. There are a relatively large number of estimates of the costs associated with a given environmental problem, eg, flooding, sedimentation, or soil erosion. There are also many studies of the economics of different land uses, and the economic benefits and costs of changing from one land use to another. There are also a growing number of studies on the economic benefits of some ecosystem services.15 Examples of all of these studies are cited in Chapters 6 and 7 of this report. Some sources cited in the previous three chapters also provide information on economic values of the physical changes described.
However, the brief for this review calls for a subset of all three types of estimates: examining the estimates of just some types of ecosystem services, resulting from just some types of land use change, in response to just some types of environmental problems. No known studies address exactly this subset of issues.
A few relevant valuation studies deal with the economic costs of soil erosion and/or benefits of erosion reduction. There are many estimates of the economic costs of flooding, both from specific flood events and national estimates for all large floods over a specified time period. However, no studies relate these costs to specific land uses.
The separation of costs of sediment and terrestrial erosion is difficult to make rigorously. This is because reducing erosion may reduce sediment yield, which may in turn reduce flood risk and damage. With care, a separation can be made; but past catchment cost-benefit analyses have generally not attempted to do so. A further general comment about the studies cited is that they are very site-specific. Mitigating erosion and sedimentation in isolated catchments may have few benefits to society, while the same erosion and sediment yield in a catchment close to population and highly valued land could be disastrous. Nationwide estimates are therefore pretty well impossible, unless more studies are done of representative catchment situations (which can then be scaled up).
Monetary value of flooding reduction
Economic costs of flooding are known to be substantial. Flooding is estimated to have cost New Zealand insurers $247 million between 1995 and 2004, excluding government compensation payments (Bicknell et al, 2004). Current government programmes for flood risk management emphasise the economic value of improving flood management strategies.
Systematic attempts to estimate the monetary value of reductions in flooding first appeared as part of flood control schemes and catchment control schemes. These schemes, implemented from the 1940s onwards, initially by the Ministry of Works and Development (MWD) and later by catchment boards, entailed substantial expenditure of government funds, so cost-benefit analyses were required by Treasury. They were sometimes undertaken by MWD staff, but more often commissioned from the then Ministry of Agriculture and Fisheries (MAF, now Ministry of Agriculture and Forestry). Many of these earlier reports have recently been sighted and several of them are still accessible.
However, in these surveys:
reductions in flood damage have been estimated ex ante, by superimposing scheme design parameters on historical information about extent of past floods
benefits appear to have been costed in proportion to postulated reductions in floodwater extent (and associated sedimentation); not by survey of affected landowners
attached monetary values pre-date the 1980s economic reforms. Costs of erosion control, values of farm production etc cannot be simply updated by applying Farm or Commodity Price Index values or similar.
For these reasons we do not consider these reports as suitable sources of data for present-day estimates of flood reduction benefit.
After 1987, central government demanded better information about costs and benefits of flood and erosion control before issuing funds. In response, agencies commissioned several studies, at catchment or regional scale, that were more rigorous in their methods. These provide more reliable, and realistic estimates of monetary benefit.
The first in this category was Ministry of Works and Development’s review of the East Coast Project (ECP) (Cowie et al, 1987). The ECP began in the late 1960s as a response to already chronic land degradation problems. It was aimed at the promotion of regional economic and social development, and the establishment of productive forests. Cost-benefit analysis was undertaken on the basis of commercial forestry returns, rather than valuation of soil conservation benefits. Such benefits had been achieved in part, but only to the degree that afforestation had proceeded (at best one-quarter of target land). Taking into account the costs of afforestation and loss of grazing production, the review concluded: because of lower effective Government subsidy rates for soil conservation, few comprehensive soil conservation programmes were likely to continue in the region.
The current East Coast Forestry Project is the ECP’s continuation, as a response to Cyclone Bola (in 1988) and the worsening erosion problem on the East Coast. It has since been reviewed at five-year intervals. In the first review (1998), a very simple analysis was made of the benefits of a reduction in flood risk (hence flooding costs avoided), using the recalculated cost of the Cyclone Bola disaster. The flooding benefits from afforestation (assuming all target land being planted) were estimated to lie between $1 million and $18 million. However, this estimate was crudely done and the calculations were more to show the magnitude of possible benefits than to portray accurate values.
In the Waikato, Project Watershed was established by Environment Waikato in 2000 to deal with ongoing soil erosion, river management and flood protection in the Waikato and Waipa catchments. Interventions include riparian fencing and planting and planting of other erosion-prone slopes. Some monitoring of the effectiveness of these actions has been initiated (B Peploe, Environment Waikato, personal communication). No cost-benefit analysis has yet formed part of this monitoring, but is likely to be included in a review of the project to be undertaken.
In the Manawatu-Wanganui region on 15/16 February 2004, a rainstorm varying between 150 and 200 mm affected many parts of the region. It caused 62,000 landslides over an area of c. 10,000 km2 (Dymond et al, 2006). MAF surveys summarised by Trafford (2004) estimate the undifferentiated costs of damage from landsliding, flooding, and siltation to be $170 million.
Monetary value of changes to low flows
We are not aware of any literature directly valuing the cost of reduced low flows through afforestation / reversion on downstream land uses in New Zealand. Such effects have been the source of some policy debates in South Island regions, but our impression is that they would not apply to large areas of afforestation, nationally.
Monetary value of sediment reduction
Some relevant information was collected by catchment boards, regional councils, or MAF, in the course of several ad hoc responses to storms during the 1980s to early 1990s. These were reviewed by Clough and Hicks (1993), and a summary of the more useful data appears in DL Hicks (1995: table 4), copied below as Table 11.
We consider the estimates in Table 11 to be reliable. They were selected because they can be under-pinned by real costs and returns, collected for the area where an investigation was carried out. Estimates from other publications or unpublished reports were rejected, as they appeared to be extrapolated from generalised farm costs and returns, civil cost estimates etc.
However, there are several disadvantages in applying the Table 11 estimates to carbon-sequestration-related sediment reduction:
they are now 12+ years old, and few in number
sediment-specific components can only be isolated by going back to source documents
it is unclear how sediment-specific components could be attached to similar terrain in other parts of the country.
Table 11: Monetary value of damage repair costs (averaged over farm area; sediment, flooding, and erosion damage combined)*
Nature of repair
Note: Costs date 1988–1992.
* Source: DL Hicks (1995: table 4).
Some extra information is contained in a valuation of the benefits and costs of soil conservation to the Bay of Plenty region (Weber et al, 1992). The study was conducted by a mixture of direct costs analysis and a survey of willingness to pay for soil conservation benefits. The benefits and costs of all types of soil conservation (not just afforestation) were evaluated and ecosystem service benefits other than soil conservation investigated. Soil erosion was addressed as a problem specifically in regard to its effects on water quality. This is tantamount to a cost-benefit analysis of soil conservation works, in terms of their value for reducing sediment and nutrients in receiving waters. The report concluded that soil conservation was a worthwhile activity from an economic perspective, with a net present value of at least $2.7 million to the region at an internal rate of return of 11.3%.16 This study did not isolate flood and erosion reduction benefits.
Monetary value of terrestrial erosion reduction
There is an extensive literature on the subject of benefits from soil conservation. The first published evidence from New Zealand was bulletins and pamphlets produced by the Soil Conservation and Rivers Control Council from 1945 to 1973. These promoted diverse soil conservation practices: burning control, pasture oversowing, aerial topdressing, tree planting, runoff diversion, gully repair. They cited reduced erosion and increased farm production, on the soil conservation reserves and experimental farms where these techniques were trialled. Clough and Hicks (1993) reviewed these early publications, considered that they contained anecdotal descriptions of perceived benefits that may well have been real, but lacked rigorous measurements to which monetary values might be attached.
Establishment of the Water and Soil Science Centres in 1974 led to another series of miscellaneous publications and publicity leaflets. Some of these reported scientifically measured effects of various soil conservation practices eg, reduced erosion, improved soil properties, better vegetation cover, pasture and timber yields. Their papers in scientific journals discussed the measurements and sometimes attached estimates of their monetary value, but ironically were not accessible to the farming community. A ‘plain language’ compilation of information from these documents has been published by MAF (DL Hicks, 1995). Its summary of the reductions in crop and pasture yield due to erosion is reproduced below as Table 12.
Table 12: Range of percentage reductions in crop and pasture yield due to erosion*
Minimum reduction (%)
Maximum reduction (%)
Surface erosion, cropland
Surface erosion, pasture
Surface erosion, tussock
Deep mass movement, pasture (initial)
Deep mass movement, pasture (re-grassed)
Shallow mass movement, pasture (initial)
Shallow mass movement, pasture (re-grassed)
* DL Hicks (1995: table 3).
Information about the extent to which these reductions have been reversed through soil conservation practices, is scattered through the text of Hicks’ compilation. It does not appear as a single table. Several regional council, central government, and research organisation reviews of erosion control in the period 1996–2006 have used these numbers, but do not appear to have attempted attaching monetary values.
More recently, the economic, environmental and social impacts of afforestation in the East Coast region were studied by McElwee (1998). He modelled farm-scale and regional effects of full afforestation with pine and space planting with poplars. At the farm scale, both pine and poplar forestry had a greater Net Present Value than pastoral farming. At the regional scale, his results indicated the same order of NPV relative returns of the three land uses. He noted further implications in terms of environmental and social values at the regional scale:
Overall, therefore, it appears that planting large stations [compared to smaller farms which are more profitable to individual investors at the farm scale] would maximise employment and soil protection, and may also maximise NPV ... Thus by encouraging forestry development on the severely eroding land classes which are typical for large stations, the East Coast Forestry Project may well be helping to maximise employment and the financial return to the East Coast regions as well as achieving the stated goal of controlling soil erosion.
All work mentioned so far has looked at local costs and benefits. Ericksen (1984) attempted the first nationwide estimates of flooding costs. His review contains some aggregated data for major floods between 1960 and 1984. It appears reliable, being based on insurance claims for property damage, production losses from farmland, and repair costs for river protection works. However the review needs updating to present-day dollar values.
A study by Krausse et al (2001) is the first attempt at estimating national economic costs of soil erosion and sedimentation in New Zealand. The authors use a framework modified from earlier work by Clough and Hicks (1992), as well as American approaches.
Krausse et al (2001) estimate average annual cost to be $127 million (for soil erosion and sedimentation combined), but caution (with commendable honesty) that their estimate is order-of-magnitude ie, the true figure could be anywhere between $12.7 million and $1270 million. However, because the authors have attempted to err on the conservative side when estimating, their true value is likely to be higher than the mean estimate of $127 million. Furthermore, the highest component of costs is lost agricultural production, estimated at $37 million annually. They note that predicted changes in climate patterns for New Zealand include increased storminess in many regions and drier conditions in eastern New Zealand. Both trends would increase lost agricultural production from mass movement in hill country and surface erosion, respectively.
This study should be compared with an earlier estimate of an annual cost of erosion of $30 million for rural land and $3 million for urban land (Hawley, 1984, quoted in Glade and Crozier, 1996). This figure is almost certainly an underestimate: for example it gives a direct cost of only $1.52 million for the on-site cost of erosion in Cyclone Bola, compared with the usually quoted cost of more than $100 million for the total costs of the damage caused in that storm.17
The Krausse et al (2001) study, although still the best national estimate of the costs of soil erosion, is simply an aggregation of data on different costs. It does not attempt the type of first-principles approach summarised at the beginning of this chapter, and its range of costs spans two orders of magnitude. The authors discuss why they were unable to produce a better estimate.
national agencies (the former MWD, MfE and MAF) and regional agencies (catchment boards and regional councils) have never systematically collected or collated data about storm damage or costs for more than a few years at a time
costs of erosion and flood damage are known only for a few events that have been well-documented eg, Cyclone Bola; these are too few to be a reliable basis for making nation-wide estimates.
Finally, there is an important qualification to any estimate of costs of soil erosion: they cannot be directly correlated to benefits from afforestation / reversion. Some costs of soil erosion are not relevant to afforestation / reversion, and in many other categories the costs and benefits are not equivalent. For example, the large loss of pasture production value associated with hill-country mass-movement erosion would represent a cost, not a benefit of afforestation or reversion – because afforestation / reversion would take the land out of pastoral production.
Conclusion: monetary estimates
It is clear from the works cited above, that there is plenty of information available on local benefits. Some of it has already been extrapolated to nationwide estimates, but these are either outdated, or have large uncertainties. There may be ways to get around the absence of reliable nationwide information, for instance by using regional information, where these are of sufficient quality and resolution.
The studies discussed mainly looked at total costs (not the change in the costs due to an increase in areas planted). Only a few of the studies actually tried to calculate the benefits of some of the changes in ecosystem services (loss of damages, water quality improvement).
Quantification of those benefits at nationwide scale will require new information on areas to be afforested and reverted. For greater certainty, they will also need updated information about reductions in erosion and sediment yield, changes in flood risk and magnitude, changes in damages, and changes in ecosystem services (direct and indirect).
These possible methods are discussed in the next chapter.
14 See also Krausse et al 2001 for a detailed classification of erosion and sedimentation effects.
15 These include Patterson and Cole (1999), Ensis and NZIER (2006a, b), Morten (2006), and Butcher (2006). See also http://www.doc.govt.nz/templates/MultipageDocumentPage.aspx?id=40121 for a summary of recent Department of Conservation work on the value of ecosystem services provided by public conservation lands.
16 As with many regional council valuation studies, this report focussed heavily on the relationship of private to public benefits, in order to provide a basis for rating.
17 Although this review includes only New Zealand material, it could be noted that the Glade and Crozier (1996) review cites a yearly cost of direct damage from landslides of NZ$243–743million in Japan. This estimate is relevant as Japan’s area is roughly comparable to New Zealand’s (370,000 km2 and 270,000 km2 respectively) and has similarities in terrain and rainfall. The Japanese estimate is likely to comprise mainly infrastructural damage and not the costs of lost agricultural production.