\section{Discussion} This section discusses common problems of EIA implementations as they relate to the RMA, as well as selected issues that are specific to New Zealand's integrated and devolved approach to environmental assessment. \subsection{The quality of assessments} Since the RMA does not prescribe a specific process that ought to be followed in preparing and reviewing an AEE, the quality of assessments and the efficacy of their review through the councils varies greatly. Furthermore, due to the very broad definitions of `environment' and `effects' that the RMA adopted, a wide range of projects falls into the set of proposals that require assessment, creating an enormous volume of assessment and review work \parencite{practitioners}. % TODO: consideration of alternatives: how about the RMA? According to a survey of EIA practitioners \parencite{practitioners}, the lack of central guidance on impact assessment practise makes it difficult for the assessors---planning professionals and engineers who are often minimally trained in EIA---to produce adequate assessments. The guideline presented by the Fourth Schedule of the RMA is often overvalued as an issues checklist for assessment. As a result, assessments are not seen as a means to enable affected parties and decision makers to find a well-informed compromise that is acceptable by all participants, although aiding decision-making processes is a core principle of EIA; instead, a majority of survey participants primarily aimed to fulfil the requirements of the Fourth Schedule in preparing an AEE and was not concerned with following international EIA best practise \parencite{practitioners}. %- failure to predict important impacts A review of consent processing performance by the Ministry for the Environment further revealed that councils rarely reject subpar resource consent applications as permitted by section 88(3) of the RMA; much more often faulty applications are accepted and gradually improved through requests for additional information in line with section 92 of the RMA \parencite{performance}. It is doubtful whether poor quality assessments significantly improve through this course of action. It is clear, however, that this approach not only delays the processing of resource consents, but also increases the likelihood of poor quality applications slipping through. According to \textcite{retrospect}, ``EIA generally continues to bring about only relatively modest adjustments of development proposals.'' Given the tendency of councils to accept subpar assessment reports and the fact that only a little more than half a percent of all resource consent applications are eventually declined \parencite{rma-survey}, and considering that AEE practitioners rarely employ more advanced means of assessing impacts than overly simplistic checklists \parencite{practitioners}, it seems very likely that this statement applies to assessments in resource consent applications as well. \subsection{Participation of the public} Although the use of objective measurements and scientific methodology is considered EIA best practise \parencite{principles}, EIA is neither science nor is it an objective process. As environmental impact statements are produced by project proponents with the goal to convince decision-makers of the benefits of the project in question, the report is a subjective statement or even a piece of project advocacy \parencite{beattie}. In recognition of this inherent bias, the EIA process calls for the participation of the general public, in particular the participation of affected individuals or interest groups \parencite{wilkins}. It is therefore rather disappointing that even in recent reviews of international EIA practise, public participation remains on a fairly low level \parencite{eia-state-of-the-art}. Some of the main barriers to public participation cited by \textcite{eia-state-of-the-art} are: poor knowledge of the public about the process; poor provision of information; failure to influence the decision-making process; poor execution of participation methods; and regulatory constraints. While councils usually engange the public during the consultation phases of the plan formation process, the picture on the resource consent level is a different one. According to the 2010/11 survey of local authorities the New Zealand \textcite{rma-survey} carries out every two years, only about six per cent of all resource consents in the two-year period were notified in some way, with only four per cent being publicly notified (``poor provision of information''). Hence, although the public can influence the framework relative to which resource consents are evaluated, there is limited opportunity for the public to affect the outcome of the actual decision-making process; this situation may result in reduced willingness to participate in areas where public participation is still possible (``failure to influence the decision-making process''). % screening is political because it depends on the values of those % who perform the screening \subsection{Cumulative effects and the devolved mandate} What sets apart New Zealand's approach to environmental assessment from those of other countries is the devolved mandate. The distribution of responsibilities to the local levels of government, however, brings about difficulties in effective environmental management. Project-level EIA usually does not address cumulative effects well, i.e. individual minor effects of several projects that result in serious impacts when combined, because this would require regulation and monitoring at a higher level \parencite{eia-state-of-the-art}. When resource consent applications are processed independently from one another at the local level, their aggregate cumulative effects are easily overlooked. Although the RMA specifically includes cumulative effects in the definition of effects that have to be considered (Section 3), it is still up to the council to review an AEE with regards to cumulative effects. The quality of this review crucially depends on the experience and the resources available at the local level to scrutinise an AEE that may not properly address cumulative effects \parencite[p 267]{furuseth}. A joint hearing process has been used in the past to successfully overcome this limitation for individual projects that require multiple resources consent applications to be considered \parencite{fookes}. %- poor communication The same problem exists for `Permitted Activities' whose impact is considered too minor to warrant an assessment of effects. The RMA does not demand an assessment of the cumulative impacts of `Permitted Activities'. According to the RMA Survey 2010/2011, a surprisingly large percentage of regional councils (91\%) carried out monitoring procedures and reported on `Permitted Activities' for the sake of assessing cumulative impacts \parencite{rma-survey}. According to the same survey, however, only 68 per cent of those activities that required both resource consents and monitoring were monitored by regional and territorial councils. As a result, it is difficult to evaluate the accuracy of the predictions of a considerable number of AEE and the effectiveness of local plans and policies \parencite[compare][p 49]{sadler}. The effectiveness of monitoring to anticipate cumulative effects also depends on the institutional framework in which it is performed. For local authorities under the RMA, national policy statements and national environmental standards are supposed to provide reference points for local plans and policies that determine the `intensity' of environmental monitoring, yet the Ministry for the Environment has been relatively slow in publishing these national guidelines \parencite{miller2010implementing}. Although according to \textcite{sadler}, the integrated approach to EIA encouraged by the RMA should, in theory, be sufficient to establish a ``context and parameters for subsidiary EIAs, which are required for all resource use consents'', due to slow implementation of the RMA ``local governments still rely on project EIA rather than undertaking policy and plan-level assessments'' (p 146). The relative lack of guiding constraints on local plans favours regional differences in the implementation of environmental management practises \parencite[see case studies in][]{discussion}.