Beyond valuation. Monetary aggregates for the SEEA-EA. The Italian proposal

After years of experimentation, documents, meetings, consultations and negotiations, the SEEA EA re-ignited the debate on the monetary quantification of the value of nature. Now more than ever, there is a need for an original approach, able to align the SEEA EA to the principles of national accounting, to the relevant economic theory and to 'strong sustainability'. In this article, we outline a technically sound alternative to the currently dominant approach, namely "valuation" of ecosystem services (ESs). The basic idea is to recognise the specific meaning and usefulness of each of the numerous monetary estimates proposed for valuation, starting with those included in the SEEA EA, without forcing national accounting principles: we promote the transition from the narrow concept of "monetary value of ecosystem service" to the wider one of “monetary values connected to/dependent on ESs”. Nothing new in terms of evaluation techniques; only a simple, but rich in implications, innovation in terms of interpretation and conceptualisation of the values generated by existing monetary values estimation methods.


Italian approach
After years of experimentation, documents, meetings, consultations and negotiations, the SEEA EA (United Nations 2021) re-ignited the debate on the monetary quantification of the value of nature. Now more than ever, there is a need for an original approach, able to align the SEEA-EA to the principles of national accounting, to the relevant economic theory and to 'strong sustainability'. In this article, we outline a technically sound alternative to the currently dominant approach, namely to "valuation" of ecosystem services (ESs). We define valuation as the determination of the monetary value (EV) of ecosystems and their services. This encompasses both exchange and welfare values. In this paper, we focus, in particular, on the exchange value concept, which plays the pivotal role in the SEEA EA, as for monetary values. After quickly retracing relevant reasoning and literature ( §2), the salient features of the approach proposed by Italy are exposed. At its heart, we place the relevant monetary values that are able to capture the economic importance of ecosystems without forcing national accounting principles ( §3). The basic idea is to recognise the specific meaning and usefulness of each of the numerous monetary estimates proposed for valuation, starting with those included in the SEEA-EA, as opposed to assuming them all as "EVs of" ESs. We promote the transition from the narrow concept of "monetary value of ESs" to the wider one of "monetary values, or aggregates, connected to ESs", i.e. from the search for a single value, supposed to be representative of the economic importance of ESs, to the consideration, as valuable input to decision-making, of a range of economic cost and benefits associated with the existence or non-existence of ESs. Nothing new in terms of evaluation techniques; only a simple, but rich in implications, innovation in terms of interpretation and conceptualisation of the values generated by existing monetary values estimation methods. The starting point of this approach is the observation -in line with the SNA and the general approach of the SEEA EA itself -that a correct assessment of the EV of ESs* can only be based on the concept of resource rent (RR). However, on the one hand, this value appears useful only from the point of view of income allocation analysis and, in particular, as a measure of the distributional implications of economic ownership (direct use or appropriation in the context of productive activities) of ESs and of existing market structures and much less for a representation of the dependence of economic values on ES. ES exchange values do not capture everything and also the SEEA EA attempts to consistently recognise that the monetary values obtained are not complete with respect to the environment and highlights the importance of multiple measures including biophysical ones. To this purpose, it is therefore important to have information on the whole value of the products and produced assets that depend on ESs, as such a value is able to provide a more adequate understanding of the economic values at risk in case of loss or degradation of ESs. This approach has been fruitfully adopted in several studies ( Roxburgh et al. 2020, Salin 2021, Almeida 2022). On the other hand, many of the methods put forward in the SEEA EA to cope, through the attribution of proxy values (imputation), with the absence of observable prices for ESs and of products derived from them, provide results that do not respond to the RR concept and are, therefore, inappropriate for SNAconsistent valuation. Yet, the estimates provided by these methods have very interesting 1 (though inhomogeneous and non-additive) specific meanings ( §4). Therefore, while imputation in these cases introduces an arbitrary twist in the meaning of the estimates, the visualisation of the rich set of actually existing monetary EVs connected to or dependent upon living Nature, brought about by these methods, can inspire greater consideration for ecosystems in public decision-making without the need for these values to be confused with those of ESs themselves ( §5).

Pluralism of values versus mono-dimensionality in well-being and sustainability measurement
Although not designed for this purpose, the main indicator of the national accounting system, Gross Domestic Product (GDP), is widely misused to represent societal progress on a mono-dimensional and monetary scale. Awareness of the major constraints of GDP in this respect and of their policy consequences has given rise to a wide range of approaches, many of which are based on the idea of 'correcting' GDP: "genuine" income (in the sense of well-being or sustainability) and inclusive wealth, able to include the monetary expression of values not considered in the aggregates of national accounts, amongst which are the ecological ones. Ecological economics, on the other hand, have placed the pluralism of values amongst its conceptual foundations (Martinez-Alier et al. 1998) and state that it is not possible to trace back the erosion of the natural, nonproduced physical basis of production (and of life) to a single and all-encompassing monetary measure. Although not necessarily linked to the tradition of ecological economics* , numerous initiatives, more and less recent, adopt a substantially multidimensional vision (de Groot et al. 2006, Kosoy and Corbera 2010, Farley 2012, Kumar 2012 and go in the direction of expanding the sets of indicators for measuring wellbeing and sustainability. Amongst the most significant ones, the Sustainable Development Goals (SDGs), adopted by the United Nations in 2015, the Stiglitz-Sen-Fitoussi Report (Stiglitz et al. 2010) and, in the Italian context, the 'Equitable and Sustainable Well-being' (BES) system regularly implemented by Istat since 2013. While supporting a powerful biophysical analysis, when it comes to monetary values, the SEEA EA implements a monodimensional approach, by looking for the economic value of ESs, homogeneous to that of commodities.

Monetary valuation and consumption of nature as capital
The fact that nature, artificial capital and other forms of wealth (social, human, institutional, cultural, spiritual) are all called "capital" does not necessarily make them substitutes and, obviously, they are not in the real world even when they are all valued according to the same monetary meter. The assumption of substitutability, on the other hand, is rooted in neoclassical economics (Solow 1974, Hartwick 1977, Solow 1986, Pezzey and Toman 2002, Arrow et al. 2004, for which other forms of 'capital' can be substituted for 'natural capital'. This reduces the sustainable management of environmental resources to that of aseptic financial investments, in the context of models that lead to decisions regarding the extraction and consumption of resources on the sole basis of economic convenience. Although formally extended to a wide time horizon, they assume the perspective of a given moment in time (the present). These models do not contemplate decision-making processes similar to the real ones, which are based on the consideration of a plurality of values, but apply the exclusive criterion of maximizing utility, regardless of the actual ecological status and the specific future consistency of the residual stock of natural capital. Therefore, such abstract models cannot provide correct guidelines for environmental policies concerning the future or large-scale ecosystem changes. By focussing on value in exchange (i.e. on substitutability in economic agents' utility functions), the SEEA EA, when it comes to economic monetary values, clearly embodies a weak sustainability perspective. We deem that it should also be able to connect to the rich political economy tradition, according to which it is more useful and robust for the sustainability of the economic process and long-term social well-being, to constrain the objective of maximising economic convenience in the use of nature and its services to non-declining stocks and ecological quality of nature over time.

Market or institutional failure?
Failure to recognise the benefits deriving from ecosystems and the costs deriving from their loss is not simply a market failure, whether this is strictly understood as the nonexistence of markets for natural public goods accessible to all and free of charge or as imperfection of existing markets: it is also a broader institutional failure. Many of our institutions have proved unsuitable for managing the social costs of economic activity and unable to rationalise access to natural public goods to the necessary extent. Policies often aggravate the problem by subsidising people and businesses more to exploit nature than to protect it, by not limiting access to essential resources that should not be compromised and by giving priority in the allocation of public resources to unsustainable economic activities. A conservative estimate of the global total cost (Dasgupta 2021) of subsidies that harm nature is around $4-6 trillion annually. Institutional arrangements (IAs), fit to protect global public goods, are lacking. Nature needs to enter economic and financial decisionmaking and, to do so, it is useful to broaden economic measurements in a multidimensional direction. In fact, in the face of significant risks and uncertainties about the consequences of ecosystem degradation, economic rationality itself suggests the preference for quantitative restrictions in use rather than mechanisms of pricing the use of nature itself. After all, it is almost obvious that it is cheaper to preserve nature than to restore it once it has been damaged or degraded, assuming that this is possible. By relying exclusively on the results of price formation dynamics that reflect the scarcity of the resource, there is no a priori guarantee that goods essential to survival are not consumed beyond their natural regeneration capacity anyway, nor that access to essential resources does not occur mainly on the basis of wealth (i.e. with scarce ESs becoming luxury goods).
In this regard, it should be remembered that politics and institutions can decide -as in fact they often do -to artificially create the scarcity of the resources to be protected by establishing appropriate IAs. This can be done not just by creating markets, but also by regulating access to resources with quota mechanisms, which seems appropriate, especially for ESs essential to human existence.

National accounting and ecosystem services (ESs) Centrality and meaning of exchange value (EV)
The SEEA EA aims at consistency with SNA principles and concepts. The SNA is all built around actual EVs, i.e. the "values at which goods, services, labour or assets are in fact exchanged or else could be exchanged for cash" (SNA, 3.118) under the current (and not some hypothetical alternative) IAs (SNA 3.119; SEEA EA 8.15,9.30). We adhere to the view that EVs can only arise in voluntary transactions, i.e. between willing parties (the expression "willing sellers and buyers" is recurrent in the SNA, for example, 3.199). In this view, when EVs are concerned, reference cannot be made to something that nature and humans exchange, i.e. to ESs as such, but necessarily to something that may circulate within the economy, i.e. between economic agents, namely the right to use ESs. Recognising the social relationship nature of this exchange, whether actual or hypothetical, is crucial to properly understanding the meaning of valuation and of its results. We are aware of the existence -and even of the dominance -of different views, dealing with the ecosystem as if it were an autonomous agent able to take decisions of its own or assuming that exchange value can arise within a single agent (in its direct relationship with nature) , but we trust that the illogicality of such a position (no matter how widely accepted) emerges to anyone who is open-minded to "philosophical" (basic epistemologic, deeply scientific) issues such as the pre-existence of the social institution "market" as a precondition for the existence of whatever exchange value, the difference between market/exchange value and welfare values and the importance -for the very arising of exchange value -of the two agents of the Edgeworth box being sentient (if not human) beings with a utility function and decisional autonomy.

Benefits and valuation
"ESs are the contributions of ecosystems to the benefits that are used in economic and other human activity" (6.9 SEEA EA). "Benefits are classified as either SNA benefits or non-SNA benefits." (6.17).

SNA Benefits
According to SEEA EA 6.17, ESs are "inputs into an existing […] joint production process", from which SNA benefits stem. The EVs of these partial contributions of ESs to output are already included in the NAs, hidden in produced goods and services' EVs. They can only be highlighted as the share of these EVs that is appropriated by their economic owners (see SNA 3.26), that is, as rents corresponding to these resources' control. Valuation methods that do not provide estimates of this share do not lead to the quantification of the EV of ESs, but to something else, depending on the method applied (this will be dealt with in a subsequent section). The implications, in terms of actual transactions and market prices, of the relevant IAs -such as the existence or non-existence of ESs markets, of subsidies for their provision or of taxes on their use -are already embodied within NAs and only need highlighting. It should be noted, moreover, that SNA benefits are much more diffused than suggested by the distinction. In particular, SNA benefits are everywhere in real estate services markets: as the hedonic cost method suggests, a nice view or proximity to amenities or higher protection thanks to flood control and soil retention usually result in higher renting prices. Of course, this contribution to monetizable benefits cannot be always easily estimated, but this does not mean it does not exist.

Non-SNA benefits
Non-SNA benefits, defined in SEEA EA 6.18, are not included in the production boundary of the SNA. This basically means that no income (rent) is recorded in national accounts, neither as such, nor implicitly in the value of transactions that must be recorded according to the SNA. No production activity -as defined in the SNA -happens here. Production activities as included in the usual NA SUTs are not involved. ESs are "used and enjoyed by people and society" directly, with no mediation of production activity. Even if the SEEA EA does not explicitly state it (though the draft version of chapter 8 did), the production boundary is expanded as to include people's enjoyment of cultural and regulation services. These activities are supposed not to generate output in SNA sense and, therefore, the ESs involved do not generate monetary rents for those who use them. This can be because these ESa are not perceived as scarce and, therefore, not marketed or because they simply cannot be exchanged and the current IAs do not allow appropriation of the preconditions of their enjoyment, so that the consumer surplus remains all with the user of the ES (as in the case of cultural services). Therefore, it is not possible to highlight their EV in NAs, as there is none. Nevertheless, the SEEA aims at finding their EV. This requires that some other values are imputed as if they were the ESs' EVs. These are found in produced assets and economic activities and products that would emerge or disappear in case the availability of the ES changed in either direction. All are hypothetical valuations/ imputations, of a different kind from those already admitted in the SNA. The latter imputations are based on production costs or value of similar products: none of the two concepts is applicable here. Instead, the SEEA EA admits use of such diverse values as that of the activity that would be necessary in order to restore the lost ES, or to substitute it, for certain regulation ESs; value of the trips that would not have been made if the recreation in nature opportunity were not there; value that an ES marketing activity would have, i.e. consumer surplus that could be extracted by enforcing property rights on the conditions of access to an ES (this is the simulated exchange value case; Caparrós et al. 2017). Finally, the SEEA EA typically assumes, for non-SNA benefits, that they are "extracted" and implicitly exchanged by users with themselves and often by governments as trustees on behalf of society at large. This is a self-justifying assumption, giving rise to self-balanced accounting items.

Policies determine institutional arrangements (IAs) and these determine prices
It is important to recognise that the prevailing IAs (property right regimes, laws and regulations, cultural approaches and customs etc.) determine both the EVs that can be observed in actual transactions and those that can be estimated for imputation purposes. ESs scarcity, on which their prices depend, may itself be 'natural' or 'artificial'. In the latter case, it is due to restrictions to access imposed by public or private control. The rent that can be derived by controlling ES is, in general, not connected to their real economic importance, nor to their natural scarcity, but to the "institutional arrangements surrounding the use of the ecosystem" (SEEA EA 9.37). NAs take the current IA, 'as-is', as the reference IA, from a neutral perspective, refraining from all judgement on it or on the ideal context for the measurement of EVs. The prices used in NAs reflect the current IAs, regardless of all possible bias given by externalities, legal and regulatory dispositions and market imperfections, including rents from dominant position. If it were not so, the NAs would be a normative instrument and not a cognitive one. Such 'agnosticism' extends from the existence of markets to their competitive structure (perfect competition, monopoly, oligopoly, protected or contestable, subsidised or taxed…) and price formation mechanisms (SNA 3.119). This absence of preference raises the issue, for any valuation diverging from the rent calculation, of which market structure is implicit in the choice of the method or should be adopted when expressly required (as in the SEV method). This problem may also arise in the SNA context, but in the case of goods produced and/or for which market exist, in general, makes the issue much less relevant, with non-hypothetical transactions in actual markets (of inputs or similar products) supporting valuation. It can also be noted that, in the most important case of SNA imputation -GG output -valuation of the output is necessary in order to include public employees wages in VA (without which, SUTs would not be balanced).

The EV of ESs as a distributional issue
SNA-consistent valuation, i.e. assessment of the rent commanded through ESs control, does not allow to capture ESs' generation costs, nor their economic importance from a social point of view, nor the dependence of economies from them, but only their 'particular' EV. In this sense, valuation is bound to fail vis-a-vis its main rationale, i.e. demonstrating nature's economic value. Again, the specificity of ESs' with respect to other assets' services, follows from their being non-produced, which implies that only scarcity and IAs play a role in determining their rents. The under-representation of a produced asset "real value" in the SNA may be dramatic, but in the case of ecosystems, it is fatal, as it concerns something that cannot even be reproduced.
The 'particular' EV of ESs is determined -beyond their usefulness -by the perceptions that able-to-pay economic agents have about scarcity and, therefore, ultimately by IAs, market structure, demand elasticity. Rent is the accounting reflection of a social relationship established by property rights on ESs, which contributes to the wider income distribution situation. Its measure only provides information about the allocation of income amongst economic agents, namely on the EV that can be subtracted to other production factors, thanks to the negotiating power provided by exclusive access to and 'extraction' of ESs* . The dependence of the measure of nature's contribution from the almost arbitrary appropriation of a share of added value by 'landlords', poses problems that trespass into the ethical field, in the sense that rents are the result of a negotiating 'confrontation' between conflicting particular interests, rather than reflect the importance of the element being valued. Again, the fact that this may apply to other prices as well does not detract from the importance of highlighting this in the specific case of nature, also considering the main rationale of valuation hinted to above.

From monetary values connected to ecosystems and their services to the value of ESs and back
After recognising the concept of (resource) rent as the reference one for an SNAconsistent valuation of non-produced resources such as ESs, its inapplicability in the case of non-SNA benefits, the fact that many of the techniques proposed in the SEEA EA do not provide estimates of the rent and are, therefore, not fit for SNA-consistent valuation (as explained above), that SNA-consistent values crucially depend on IAs and only provide information about income allocation, we need an alternative for properly highlighting not just the EVs of ESs, but also their wider economic importance. To this end, we put at the centre of the stage the specific meaning, information value and usefulness for decisionmaking of each of the numerous estimates, offered by the different methods put forward for valuation.This only requires resisting the 'imputation' temptation, i.e. to the arbitrary assumption that all of these values express a generic, homogeneous, additive and direct measure of ESs value. Indeed, decision-makers need a range of information on the potential effects of the different policies, including on how changing the institutional context may influence the nature (monetary/non) and entity of the values at stake.

ESs' own EVs and other connected EVs
The pivotal role in the Italian proposal is played by the difference between the concept of "monetary value of ecosystems and their services" and that of "monetary values connected to (or relevant for) ecosystems and their services". The former expression points to "the EV that a given ES has or would have if it was on the market". The latter expression refers -besides obviously to resource rents as actual EVs of ESs -to all the observed or estimated values of actual or potential transactions, that somehow depend upon, or are linked to, that particular service. For instance, expenses for nature tourism crucially depend on nature's ability to provide valuable outdoor recreation services, but they do not represent the exchange value of the ES deriving from this ability. These expenses are determined -besides, of course, by attraction -by tourists' spending ability (income) and preferences and by the prevailing conditions in markets, such as travel and lodging, which, in turn, depend on external factors. Changes in the used quantities and their monetary values may reflect changes in any of the involved markets, having nothing to do with the quality of and ecosystems' ability to provide the ES in question. Similarly, flood protection services' use depends primarily on human presence in the area at flooding 3 risk, so that their increase is usually not good news, as it mostly depends on increased demand -land consumption and soil sealing -rather than on increased availability of this ES where it is needed and lacking.

A key concept: dependency
Understanding "the relationship between ecosystems and standard measures of economic activity" (SEEA EA 14.63), namely the dependence of specific activities from well-identified ES, is the best that satellite accounts can do, as for what concerns elements not recognisable, nor implicit, in the core NAs. Physical Supply-use tables for ESs show -like the corresponding tables for natural inputs in the SEEA Central Framework) -which activities or sectors depend from which ESs, how much of them they use and which ecosystems make them available. Monetary data cannot be accommodated in the same tables, as such data can only measure internal flows of the economy, that are exchanged between economic agents. However, the EVs connected to ESs are present in SNA aggregates and tables and can be highlighted within them (as an example, see Driver et al. 2021* ). Such monetary measures may concern the direct expected impact -for example, agricultural output at risk in the absence of pollination -but also the indirect ones, by measuring the multiplicative economic effects of any given hypothetical change in ES provision. No matter how low the total EV of ESs may appear according to SNA-consistent valuation, economy's dependence upon them -the existing EVs at risk -will always be very high. Additionally, this also provides a possible explanation of the significantly different orders of magnitude between the estimates given by the various methods. The following hypothetical situations provide some examples of dependency (or "connection"): the estimated value may be that of actually existing economic flows or stocks, recognised in the NAs, which would disappear if the ES disappeared, for example, ESs embodied in products such as crops or that would disappear if the ES were to appear, for example, when trees newly planted in an urban area start providing air filtration services, so that the demand for health care services diminishes. In both these cases, the ES lost or coming into existence is connected to economic activity and value by the fact that ESs are necessary inputs for specific economic activities or their presence or absence influences the value of existing economic activities.

Monetary values connected to ESs provided by SEEA EA valuation methods
Monetary aggregates connected to ecosystems and their services that are actual flows are always included in NAs and often made explicit in monetary environmental activity accounts (SEEA CF chapter IV). Other monetary aggregates connected to ecosystems and their services are hypothetical flows, for example, losses or additional costs, if the service is replaceable, which would derive from losing ESs or from having more of them. Although the SEEA EA has a ranking of preference for the estimation methods it puts forward, at the top of which feature those that we consider fit to represent actual EVs of ESs, it considers them all able to express this concept. We will now examine these methods, looking for a more straightforward interpretation of the information they provide than that required by imputation and homogenisation of the meanings to the single indistinct concept of EV of ESs. The different interpretation has implications for the use in decision-making. These mostly are straightforward. We concentrate here on the meanings, leaving to further developments of the approach the explicitation of the links with policy use, for whichdifferently from the SEEA EA -we do not have a priori preference for a specific concept (and, if any, not for the EV when it comes to representing the economic importance of ESs). It is sufficient for the moment to underline that the link with the discussion under "theoretical premises" above is given by the non-general-additivity of the values.

Actual EVs of ESs
One important distinction here is between: a) ESs as such (non-produced physical contributions to benefits); b) the right to use "a)" (to which transactions may refer); c) ES as (part of) a product, i.e. good and services incorporating "a)". The latter embraces a range that covers all products, going from those in which the human contribution is hardly identifiable, the name of the product is that of the ES itself and the difference with "b)" is doubtful, to those where the presence of "a)" is so little that it is usually not recognised and is hardly quantifiable.

Specific observable EVs of ESs
In the SEEA EA, this method -considered "the most direct method for measuring prices and estimating values for the accounts" -is introduced by way of examples. Let us take the first one: "for example, if a wetland provides services of water purification and the owners or managers of that wetland are able to charge the water company that abstracts the water for municipal uses, there is transaction in ESs provided by the ecosystem that can be recorded" (9.28). We are here near the first end of type "c)" casuistry. The transaction is between two economic units. Owners or managers of that wetland -not the ecosystem -provide the action or inaction that is necessary to keep an essential feature (purity) of the water used by the company. The observed price is that of a derived product -"water purification services of ecosystems" -which does not yet feature in official product classifications, but may do soon. From this price, to get one representing the specific EV of this ES, i.e. of "b)", all other input costs, such as transaction costs (e.g. services of lawyers for writing and enforcing contracts) or protection costs (e.g. to avoid the wetland being used as a dump by third parties or that it becomes eutrophic) have to be subtracted. A case responding to type "b)" is that of "observed prices from emission trading systems which may be used to estimate prices for global climate regulation services based on carbon retention" (SEEA EA 9.32). Prices, in this case, are observed only for the 'retained carbon' quantities for which a corresponding emission right is actually sold (as it is supposed to offset those emissions), while if applied to quantities not sold, the case is that of 'EVs of similar ESs', dealt with in the following section.

EVs of similar ESs that are traded
This method is foreseen, for products, in SNA 3.123 and it can be applied in cases where, for a same ES, in different locations, situations co-exist where prices are observable and where no market/no price exists. Clearly, we have two different institutional contexts here and the prices will reflect -as the SEEA EA itself states -"the existing institutional context" (9.34). Of course, this means 'existing in the other context, where the market is the societally chosen regulation mechanism for access to the ES. This applies also to the implications for income distribution. The method provides figures that represent the income shift in favour of the owners, should the latter be able to impose the same market conditions as those prevailing for similar services elsewhere. Such estimates are reliable, according to the SEEA, if "the flows of (non-marketed) ecosystem services […] are not significant enough such that they would alter the observed price of, and demand for, the good or service". A case where flows are surely significant enough to make the estimates unreliable is that of carbon retention services, where the method is widely applied by transferring observed prices from emission trading schemes to non-traded quantities of this service.

EVs of ESs embodied in market transactions
The methods "residual value and resource rent", "productivity change" and "hedonic pricing" represent different ways to isolate the part of EVs of goods or services embodying ESs that is specifically attributable to the control of ESs, as a differential income going to the seller of those goods and services. The RR, in particular, is obtained by "deducting the cost of all other inputs, including labour, produced assets and intermediate inputs". The concept at the basis of all these methods refers to a quite general situation: no marketed goods could ever be realised without some ES and, vice versa, all ESs are inputs in the production of some marketed product. These valuations, consistent with the SNA, provide, as discussed, a poor measure of what is at stake, as they are related neither to the ecological value of ESs, nor to their social value, but represent only the income appropriated by ESs' economic owners, i.e. by those who use them in production or benefit in asset property. Of the three methods under the present heading, the first two are especially fit for the valuation of services, such as those of water and other ecosystem inputs -for example, pollination -in agriculture (Capriolo et al. 2020), while the third one can be used to know how much of the real estate income can be considered an economic benefit for the owner deriving from the presence of ESs.

Cost of averting behaviour
This method, also called "defensive expenditure", considers expenditure directed to prevent or mitigate the subsequent retroactions ("negative effects and damages") on the social system, "for example, in relation to incurring costs associated with extra filtration for purifying polluted water, air conditioning for avoiding polluted air and so forth" (SEEA EA 9.45). The specific meaning of the aggregates provided by these estimates is clear enough as they measure the economic value of activities connected to the absence of ESs, i.e. costs to society, perhaps avoidable through ESs restoration. When applied to determine hypothetical prices of existing ESs, they provide information about the additional economic cost of coping with their possible loss.

Travel expenditure
The expenditure incurred by households or individuals to reach and access a recreational area has by itself the meaning of how much the economy depends on certain ESs, i.e. what would be lost in economic terms if the ES were lost (or if it were impossible to use it, such as, for example, due to restrictions during the current pandemic); therefore, "associated expenses" and not "ES's EV". The existence and enjoyment of ecosystems is the very reason of the production and consumption activities involved in travelling, not an input to production, nor can enjoyment be equated to production. If the recreation activity were to be dealt with as a production activity, all travel costs should be considered as production costs (intermediate inputs). The value added of the activity would be the consumer's "net enjoyment" and this surplus would still have to be quantified independently from the costs. Such an objective quantification, however, is prevented by the fact that personal experiences lack the exchangeability for cash prerequisite of EV.

Expected expenditure
The methods presented in the SEEA EA under this title include those "based on estimating the expenditure that would be expected to be made if the ES were no longer provided or were, in fact, sold on a market" (9.49). The relation to our framing of monetary values connected to ESs is direct and explicit.

Replacement cost
Knowing "the cost of replacing the ES by something that provides the same contribution to benefits" (SEEA EA 9.50), is per se surely useful for policy. In this perspective, it is crucial to understand the term "same" as referred to the qualitative dimension of ecological functions, including their localisation and not to a generic contribution to benefits. It is also important to highlight the distinction between costs for final consumption, for intermediate inputs (such as "sorghum substituting for non-priced forage in the case of a rangeland grazing ecosystem services", 9.50) and for fixed capital (as in the water treatment example)* .

Avoided damage
Knowing "the costs of the damages that would occur due to the loss of these services" (9.52) also has obvious utility for decisions, independently from the identification of these costs as the EV of ESs. In this case, we have no emerging activity, but only loss on both the ecological and the economic side. The quantification in EV terms, "particularly useful 5 for regulating services such as soil erosion control and flood control, air filtration, and global climate regulation services" (9.52), does not concern directly ESs, but what they protect. The connection to the social dimension is particularly important and interesting here. The quantification of the avoided damage (losses of assets and incomes) is, in fact, based on information on the physical damage expected in case of ESs' disappearance: additional dead, wounded and homeless people, destroyed buildings… Such physical information, together with that on not avoided damage, is surely not of secondary importance for policies.

Simulated EV (SEV) method
This method "estimates the price and the quantity that would prevail if the ES were to be traded in a hypothetical market". It "requires combining the information on the demand function with a supply function and an appropriate market structure (institutional context)" (SEEA EA 9.55). The arbitrariness of the hypothetical institutional context is evident, in the very simulation of the existence of a market -a quite specific institutional context itselfand then in the assumptions on the market structure (how competitive it is, i.e. how distributed the hypothetical ownership rights are assumed to be). The latter influences the estimates in a crucial way, as for example, perfect competition would mean EV = costs = null, i.e. the current situation. The simulated market may be, for instance, that of the fresh and purified air of a public park. The information provided by the method in this example would be about the expected reduction in visitors' number and the income from an entry fee that could be imposed, through which part of the currently existing consumer surplus would be extracted. If applied to ESs which supply SNA benefits, it should, in principle, give the actual rent of the current economic owner.

Restoration cost and other methods
Besides the methods discussed so far, the SEEA EA introduces "a range of valuation methods that are found in the environmental economics and ESs valuation literature" (9.56). We will not discuss them here, but for noting that -as the SEEA EA itself highlightsfor some of them, the methodological status of their results with respect to the NA exchange value concept is not fixed in general. Such an approach, particularly interesting in a non-valuation, policy-orientated, perspective on economic values connected to ESs, is the "restoration cost-based" approach (SEEA section 12.3.2). Information about the economic resources necessary "to re-establish pre-existing structure and function, including biotic integrity" (10.18) surely is important. The SEEA EA itself considers this approach an attempt to "measure the cost of degradation directly", rather than the value of ESs, "since there is no particular reason that the estimated restoration costs will align with the estimated loss of future flows of ESs" (12.41). The cost of degradation approach is a particularly interesting one in a perspective of integration of environment-connected values in the SNA (Vanoli 2017).
Other methods -as, for instance, stated preferences -give welfare values, which make them incompatible with the SNA, but not with our inclusive, multicriteria, framing of monetary aggregates connected to ESs. Their usefulness for policy decisions should be addressed on a case-by-case basis and having regard to stakeholder's involvement processes more than for the valuation results in themselves.

Conclusions and way forward
The need to extend the boundaries of economic analysis and accounting to ecosystems and their services should not lead to a cognitive distortion of the meaning of monetary values, calculated through the various estimation methods. As argued, the correct measurement of the EV of ESs from an SNA perspective, based on the RR concept, is reductive. The only information, useful for decision-making processes, it provides, concerns the appropriation of income by the economic owners of the services themselves. Where imputation is based on hypothetical markets' simulations, the assumptions on the Table 1.
Conceptual framing of the main SEEA EA methods for determining monetary values connected to ESs market structure and the degree of competitiveness on the simulated markets are able to affect the virtual price to the point of making the final result arbitrary and potentially misleading. The conceptual non-homogeneity of the various estimation methods and, therefore, in the estimated values, has also repercussions on the possibility to add up the different values of the ESs provided by the same ecosystem asset and, therefore, of determining a value for the asset itself through the net present value (NPV) method. The additivity of monetary values is preserved only within a homogeneous conceptual perimeter, whether based on EVs or on costs or on potential damage avoided -and, in the latter cases, only for the purpose of policy analysis and not for calculating the NPV of an EA. We acknowledge that data on the economic values at stake, beyond the mere EV of ES, represent an important complement to information on the extent, conditions of ecosystems and the bio-physical use of ES by economic sectors and households. They help us to understand how we depend on nature and what ecosystem resources we need to protect, in order to protect ourselves and our economic values. The Italian proposal for resolving the "outstanding methodological aspects" of the SEEA EA (UNSC 2021)* is that of a novel approach providing more solid accounting and statistical support to the economic assessment of the role of ES and policy. This approach is based, as discussed in this paper, on the dependencies of produced EVs (products and assets) on ESs and, more generally, on the connections between ESs and values, as well as on a correct identification of the relevant stocks and flows within NAs. Politics and decision-making processes should be -and mostly are -interested in the economic (and non-economic!) values at risk and the costs of protecting and maintaining nature in order to prevent and reduce these risks. These are not the same as the specific EVs of ES. Contrary to valuation, pluralistic information on the economic values at stake does not allow conceptualising nature as capital, but allows the choice of best policy options and does not oblige us to figure out what nature would be worth on markets whose non-existence is an expression of societal choice. Policy-relevant official statistics concerning income, sustainability, well-being etc. should not be biased in favour of market-like scenarios, but rather represent the current situation as truly as possible, leaving simulations to academic research and policy analysis. In this perspective, the Italian proposal points to the need for empirical research on how monetary aggregates -and those resulting from valuation, in particular -are currently used in policy-making, as well as for further theoretical research on how to best use the multiplicity of available monetary aggregates, based on the specific information of each of them.