BEFORE THE EPA CHATHAM ROCK PHOSPHATE MARINE CONSENT APPLICATION

Transcription

1 BEFORE THE EPA CHATHAM ROCK PHOSPHATE MARINE CONSENT APPLICATION IN THE MATTER of the Exclusive Economic Zone and Continental Shelf (Environmental Effects) Act 2012 AND IN THE MATTER of a decision-making committee appointed to consider a marine consent application made by Chatham Rock Phosphate Limited to undertake rock phosphate extraction on the Chatham Rise STATEMENT OF EVIDENCE OF DR GANESH NANA FOR CHATHAM ROCK PHOSPHATE LIMITED (UPDATED) Dated: 28 August 2014 Barristers & Solicitors J G A Winchester / H P Harwood Telephone: Facsimile: james.winchester@simpsongrierson.com DX SX11174 P O Box 2402 Wellington

2 CONTENTS INTRODUCTION... 3 Qualifications and experience... 3 Code of conduct... 3 Scope of Evidence... 4 OVERVIEW... 4 FOUNDATIONS... 5 VERIFICATION... 6 EXPECTED (A PRIORI) RESULTS... 7 OUTCOME... 8 VALUE OF CGE MODEL EXPERIMENTS AND THE RESULTS... 9 CONCLUSION Page 2

3 INTRODUCTION Qualifications and experience 1. My full name is Ganesh Nana. 2. I was awarded a doctorate in economics in 2000 from Victoria University of Wellington (VUW) for my thesis A Multi-Industry Computable General Equilibrium with Dynamic Investor and Consumer Behaviour. 3. I am the Chief Economist and Executive Director of Business and Economic Research Limited (BERL). 4. I have over 30 years of professional experience in economic research and advice. My particular area of expertise lies in computable general equilibrium (CGE) modelling, having originally participated in the development of the first CGE model of the New Zealand economy while I was a Research Officer at the Research Project on Economic Planning at the Economics Department of VUW during the 1980s. I have subsequently continued modifying, enhancing, developing, and applying that CGE model through positions at the School of Economics and Finance at VUW and now in my work for BERL. 5. I have worked full-time for BERL for 15 years, completing research projects and studies on regional development, the Māori economy, the impact of economic policy proposals, and commentating on wider economic trends, issues and debates. Prior to working at BERL, I was employed by VUW in various positions (tutor, researcher, and lecturer). I have also worked as a consulting economist at Oxford Economic Forecasting, England, as well as in the House of Commons operating the UK Treasury economic model and the IMF Multimod economic model. Code of Conduct 6. I confirm that I have read the Code of Conduct for expert witnesses contained in the Environment Court of New Zealand Practice Note 2011 and that I have complied with it when preparing my evidence. Other than when I state that I am relying on the advice of another person, this evidence is entirely within my area of expertise. I have not omitted to Page 3

4 consider material facts known to me that might alter or detract from the opinions that I express. Scope of evidence 7. I was asked to review the economic modelling and associated assessment of the Chatham Rock Phosphate (CRP) phosphate mining project. The scope of my review is limited to the economic modelling undertaken by NZIER using their computable general equilibrium (CGE) model of the New Zealand economy. Reflecting my area of expertise as well as the limited time available to me to review the work, I have not reviewed the economic assessment of the environmental effects. 8. I was provided with Report by the New Zealand Institute of Economic Research (NZIER) Economic assessment of Chatham Rock Phosphate dated March 2014 Review from Sapere Review of NZIER report dated 14 April 2014 Memo to EPA from NZIER Further information request, response to requests 28 to 33 dated 27 June I have also had a meeting with one of the report authors, NZIER Senior Economist Peter Clough, accompanied by his colleague Erwin Corong, to discuss and clarify matters. Additional information was provided to me in a note from NZIER dated 21 July OVERVIEW 10. As is made clear in the NZIER report, CGE models are a well-established and widely used tool for economic analysis. As is the case for all tools of analysis, it is important that the tools are used for the purposes for which they were designed and their way in which they are applied to the task is appropriate. 11. CGE models are designed to assess the impact of specific events, or projects, or shocks. That is, they are principally (and, arguably, ideally) suited to comparative exercises. These are laboratory-type experiments Page 4

5 where the modeller depicts an economy without a particular event, and then another one with a specified event. Observing the differences between the two pictures of the economy, the modeller assigns these differences to be a result of the event, or shock, postulated. 12. It needs to be recognised explicitly that CGE models are not designed to forecast of predict the future path of an economy. 13. However, as reflected in the Sapere review, it is also clear that there is considerable confusion as to the appropriate use of CGE models. This is of concern, especially where such confusion is used as a premise to describe such models as akin to a black box. While I do not suggest that CGE models should be readily understood or applied by all, suggesting their complexity is a reason to ignore their findings would be a backward step in terms of using appropriate information for decision-making. Indeed, in my view there are many items in daily use (e.g. the internal combustion engine) that many would have to similarly describe as a black box, but this does not preclude these same people deciding to use them. 14. In this context, it is my opinion that the use of a CGE model to assess the economic impact of a phosphate-mining project is an entirely appropriate use of such a tool of analysis. 15. Further, the use of an explicit modelling framework enables impacts to be estimated and the underlying train of logic to be interrogated. Assumptions associated with the framework can be similarly investigated and, in particular, those assumptions that can critically change the assessed economic impacts can be explicitly identified. FOUNDATIONS 16. The foundations of CGE models rest on neo-classical general equilibrium theory. These models are a step-up from the relatively unsophisticated input-output models based on inter-industry transactions. In particular, input-output models assume no changes in prices. 17. Where the specific events (or shocks) under investigation are sufficiently small, such an assumption may be appropriate. However, where the Page 5

6 shock is larger than small, then it is appropriate to adopt the more sophisticated framework of a CGE model that allows relative prices to change in response to the shock. 18. In line with neo-classical general equilibrium theory, CGE models carry an underlying assumption of the neutrality of money. That is, the real (or production) and money dimensions of an economy can be separated. In particular, such modelling depicts an economy in which money is neutral, in the sense that money prices do not influence demands and supplies and allocation of resources between sectors and agents. Rather, the core influencers of demand, supply, and resource allocation are relative prices. 19. Consequently, there are no money or currency market influences or interactions captured by the model. That is, the CGE model will not, nor is it designed to, predict or determine the effect on (or the effect of) nominal interest or exchange rate changes. As a result, questions as to the exchange rate effect of the shock under investigation are not appropriate unless one is referring to the impact on the real exchange rate. VERIFICATION 20. There is further confusion that is inherent in the suggestion that the models results have not been verified. Given the "laboratory experiment" nature of the model s simulations, the matter of verification is not straight forward. Verifying the results of the CGE model laboratory experiment is akin to the roading engineer, having determined a new highway would reduce road deaths by the equivalent of one life per year, then being asked to identify the one life that is saved each year. 21. More appropriately, the model that is used should conform to standard accepted frameworks consistent with its relevant theoretical base. CGE models originate from general equilibrium theory of Leon Walras, with the first CGE model being developed by Leif Johansen (A Multi-Sectoral Study of Growth, North-Holland Publishing Company Amsterdam, 1960). 22. I believe the ORANI suite of models as developed at the Centre of Policy Studies (Melbourne) are close to a template for CGE models, and their Page 6

7 application has been subject to extensive peer review, as noted by NZIER in their response of 27 June It is clear to me that the NZIER CGE model closely follows the framework of those of the ORANI suite of models. Indeed, the NZIER model is similar to that of the BERL CGE model which I am most familiar with and which is also similar to the ORANI model framework. 24. Further, the use of official Statistics New Zealand input-output tables by NZIER as the basis for the inter-industry transactions database for their model conforms to standard practice. EXPECTED (A PRIORI) RESULTS 25. Given the foundations of general equilibrium theory, CGE model simulations investigating the implications of a new economic activity being introduced to an area (e.g. CRP) would be expected to capture (or reflect) the following influences. The additional demand for the output of sectors that produce inputs (materials and equipment) required by this new activity. The additional supply of the new activity s production and its impact on purchasers of that product. The consequences of resource constraints (for factors of production labour and capital) as a result of this new activity attracting these resources. These consequences should be seen as changes in relative prices of capital and labour, which in turn impact on other the sectors. That is, changes in real returns to capital (and so the operating surplus of economic activities) and in real returns to labour (real wages). * Further, changes in real returns capital vis-a-vis labour will impact on the allocation of resources across sectors depending on the relative labour/capital intensity of each sector. Arising from these influences the model will report the impact on the real exchange rate and, consequently, the impact on the international competitiveness of NZ producers vis-a-vis those in the rest of the world. Page 7

8 * In this context the real exchange rate can be interpreted as the price of NZ produced goods and services relative to the price of those produced in the rest of the world. * Further, the change in the real exchange rate will impact on the allocation of resources across sectors depending on each sector s reliance on tradable vs. non-tradable activity, and its exposure to competition from imported products. The nominal exchange rate acts as the model s numeraire (or benchmark) for prices, and consequently does not change as a result of the modelled shock (consistent with the concept of the neutrality of money). 26. Ultimately, the CGE model is expected to quantify the relative magnitudes of these effects and so determine whether, and to what extent, the net effect on overall economic activity (and welfare) is positive or negative. OUTCOME 27. The results reported in the NZIER assessment of the CRP phosphate mining project conform to what would be expected from a CGE modelling analysis. 28. The competition for factors of production (capital and labour) raises their prices and so the costs 1 of their use. 29. This increase in costs results in tradable industries (i.e. those producing export products and/or competing against imported items) losing competitiveness and so reducing their production and sales. This is made particularly explicit in page 33 of the NZIER report, where it is stated that other exporters across New Zealand have lost $35 million of revenue. In addition, reductions in output from horticulture, kiwifruit, textiles, and other industrial chemicals sectors. 30. The reported change in the real exchange rate is consistent with these results. As expected, the real exchange rate appreciates in value, representing a loss of competitiveness by New Zealand producers. 1 As per note of 21 July, nominal wages rise 0.021% and rates of return on capital rise 0.023%, compared to the (without CRP) BAU simulation. Page 8

9 31. Similarly, industries that do benefit are those that are concentrated in nontradable activity. These industries benefit from additional demand arising from the additional income following the increase in real wages and rates of return to capital. This additional demand counters the negative impact of the increase in the costs of labour and capital that are also faced by these sectors. As a result, there are increases in output of the wholesale trade, retail trade, hospitality, and business services sectors. 32. The magnitude of the changes arising from the CGE model results are relatively small. This is best seen in the changes in prices, which are less than a tenth of one percent. Consequently, the overall net impact on the macroeconomy whether measured by GDP or welfare remains relatively small. 33. In my opinion the calculated magnitudes of the changes are consistent with prior expectations. This follows from noting that the gross output of the CRP activity is of the order of $350m 2. The consequential direct value added of $125m 3 is less than a tenth of one percent of national GDP. VALUE OF CGE MODEL EXPERIMENTS AND THE RESULTS 34. The use of the CGE model to assess the CRP phosphate mining project provides a framework for tracing the economic impacts and estimates the overall magnitude of these impacts within the limitations of the economic data available. 35. Further, the sensitivity simulations clearly identify the assumptions that could critically change the assessed economic impacts. In particular, simulation around the larger baseline market confirms that data estimates around domestic market size are relatively unimportant. 36. However, the world price of phosphate rock is a more important assumption in terms of influencing the overall economic impact. This is also signalled in the review by Sapere. The modelling framework enables a robust investigation of the importance of the world price assumption. 2 From Table 1 of the report, being total cost of $274m, plus profit of $75m. 3 From Table 1 of the report, being taxes ($46m), labour and overheads ($4m), and profit ($75m). Page 9

10 37. The NZIER memorandum of 27 June lists the summary results of a model experiment where the world price of rock phosphate is set at US$108/tonne (compared with US$184/tonne in the baseline experiment). This experiment indicates that the net impact on GDP declines to a positive $229m per annum, from the $280m in the baseline experiment. 38. Further, a world price at a level of US$108/tonne would require the cost structure of the CRP (as per Table 1 of the report) that is included within the model s framework to be modified. In particular, such a world price for rock phosphate would imply total revenue of the CRP operation of close to NZ$200m 4. Consequently, this revenue would be insufficient to cover the stated $274m total costs (excluding profit). 39. I note that the financial viability of the CRP operation depends on a variety of factors, including actual contract prices for CRP's product, transport and other costs. Clearly, the financial viability of the operation is an issue for CRP. The CGE model is not an appropriate tool to address or assess financial viability. However, any changes in anticipated cost structures should be incorporated into the CGE model s depiction of the industry s cost structure. CONCLUSION 40. The NZIER assessment of the CRP phosphate mining project using their CGE model provides informed, robust, and useful insights into the expected economic impact of the CRP operation. 41. The economic impacts are confirmed as positive in terms of GDP and welfare gains, although these impacts are relatively small. 42. The model results and accompanying analysis identifies the world price of rock phosphate as a critical assumption in the assessment. A relatively large reduction in the assumed world price reduces the gains to GDP and welfare estimated from the CGE model experiments, but these gains remain positive. 4 That is, 1.5m tonnes of rock phosphate per year, at a price of US$108/tonne and an exchange rate of US$0.80 per NZ$1. Page 10

11 43. However, the impact of a relatively large reduction in the assumed world price is likely to have consequences to the cost structure of the CRP operation (as it has been incorporated in the model framework). This consequential impact on the CRP cost structure has not been assessed or incorporated in the modelling. 44. I conclude that NZIER have used its CGE model appropriately to assess the economic impacts of the CRP phosphate mining project and its results concur with a priori expectations. The assumption around the world price of rock phosphate is identified as important to the model s results. Ganesh Nana 28 August 2014 Page 11