Brief Economic Assessment of Proposed Updates to Vehicle Exhaust Emissions Rule

Final Report 18 June 2012 Brief Economic Assessment of Proposed Updates to Vehicle Exhaust Emissions Rule Prepared for Ministry of Transport

Authorship This document was written by Fraser Colegrave. For further information, please contact Fraser at the details below: Email: fraser@covec.co.nz Web: http://www.covec.co.nz/team/fraser-colegrave DDI: (09) 916 1969 Mobile: (021) 346 553 Covec Ltd, 2012. All rights reserved. Disclaimer Although every effort has been made to ensure the accuracy of the material and the integrity of the analysis presented herein, Covec Ltd accepts no liability for any actions taken on the basis of its contents

Contents Executive Summary 1 1 Introduction 3 1.1 Context 3 1.2 Scope of this Report 3 1.3 Structure of this Report 3 2 Emission Rates for Old vs New Vehicles 4 2.1 Results of Latest NZ Research 4 3 Past and Current Emissions Rules 5 3.1 Timeline 5 3.2 Current Rules 5 4 Impacts on Historic Registrations 6 4.1 Impacts on Petrol Registrations 6 4.2 Impacts on Diesel Registrations 7 4.3 Controlling for Other Effects 8 4.4 Conclusion 13 5 Impacts on Historic Scrappage Rates 14 5.1 Understanding the Links between Registrations and Scrappage 14 5.2 Impacts on Age of Petrol Deregistrations 14 5.3 Impacts on Age of Diesel Deregistrations 15 6 Potential Impacts of Allowing a Rule Lapse 18 6.1 Discussion 18 7 Review of Wider Effects 20 7.1 Impacts on Market Activity 20 7.2 Impacts on the Value of the Existing Fleet 21 Appendix: Schedules from the 2007 Rule 23

Executive Summary Context and Scope Since 2004, the Government has been introducing rules to control the exhaust emissions of vehicles entering New Zealand s fleet. While the rules for new vehicles are due to expire at the end of 2014, those for used vehicles expire at the end of this year. To help assess whether they should be updated or allowed to lapse, this short report analyses potential impacts of the rules on the light passenger fleet. Emissions Rates for New vs Older Vehicles To set the scene, we first reviewed the local literature on emissions from light vehicles. The most recent research is contained in NZTA research report 441, titled Are the harmful emissions from New Zealand s light duty vehicle fleet improving? 1 The report above provides a detailed account of the emissions improvements associated with vehicles built to the new standards. For instance, using data collected across New Zealand, it shows that the average emissions rate for carbon monoxide is 83% lower for petrol cars built to the Japan 05 standard than for those built to the much older Japan 79E standard. Similarly, it shows that Japan 05 petrol vehicles emit 86% fewer hydrocarbons than Japan 79E vehicles, and 92% less nitrous oxide. The case for encouraging the uptake of these newer vehicles therefore seems clear. Past and Current Emissions Rules The first vehicle emissions rule (the 2003 rule) was introduced in January 2004, and replaced by another rule (the 2006 rule) three years later. The 2007 rule was introduced a further 12 months down the track at the start of 2008. Since then, there have been several updates to the underlying standards. At the time of writing, used Japanese petrol and diesel vehicles must be built to the Japan 05 standard, which effectively limits used vehicle registrations to a maximum age of 7 years. Impacts on Registrations We used detailed data from the motor vehicle register to look for any effects of the rules on the rate and composition of vehicle registrations. While the results initially suggested that there had been some impacts, particularly for diesel vehicles, we later found that other factors were principally at work. For instance, the analysis revealed a strong correlation between the rate of registrations and the NZ/Yen exchange rate. To further our investigations, we ran regressions of petrol and light diesel registration rates on the exchange rate, fuel prices, and a series of dummy variables to capture the implementation dates for different emissions rules and standards. The results confirmed that registrations for vehicles of both fuel types were sensitive to the exchange rate and their respective fuel prices. However, it also found that the effects of most rules were statistically insignificant. The most pronounced effects were related to implementation of the 2006 rule (at the start of 2007). However, this rule did not have any impact on used light vehicles, so the 1 http://www.nzta.govt.nz/resources/research/reports/441/docs/441.pdf Review of Proposed Updates to Vehicle Exhaust Emissions Rules 1

variable was likely to be picking up a range of other effects that occurred around the same time. Thus, overall, the results suggested that the rules had very little impact on the registrations rates for petrol or light diesel vehicles. Impacts on Deregistrations We also used data from the motor vehicle register to analyse any impacts on the rate and composition of vehicle deregistrations. The results suggested that the rules have had little, if any, impact on petrol deregistrations, but may have caused the average age of light diesel deregistrations to increase over time. Drilling a bit deeper, however, we found that the average age of diesel deregistrations had been increasing irrespective of the rules, with no discernible changes from one rule/standard to the next. Other Considerations The report also briefly considered a number of other issues, including the potential impacts of allowing the standards for light used vehicles to lapse. Intuitively, it finds that such effects would depend critically on the duration of any lapse. However, even a short lapse could have unwelcome effects if it led to a wholesale increase in the registration of much older, higher-polluting vehicles. A permanent lapse would be of even greater concern. In addition, the report concludes that allowing the standards to lapse even temporarily would not be ideal because: It could undermine the Ministry s credibility as an effective regulator, and thus have unfavourable reputation effects, and It could create significant resistance if/when standards are proposed for reinstatement. By contrast, continuing with the current standards sends a clear signal that they are here to stay. As a result, efforts to move to tighter standards in future may be more straightforward The report also investigated potential effects of the rules on the wider vehicle market, and on the prices of vehicles in the existing fleet. It finds no concrete evidence of effects in either regard. Review of Proposed Updates to Vehicle Exhaust Emissions Rules 2

1 Introduction 1.1 Context Since 2004, the Government has been introducing ever-tighter rules to control the exhaust emissions of New Zealand s vehicle fleet. The most recent rule - the Land Transport Rule: Vehicle Exhaust Emissions 2007 was introduced in 2008, and provides for the staged introduction of various emissions standards for new and used vehicles, both light and heavy. As shown in the appendix, due to the way that the rule was drafted, emissions standards for used vehicles effectively lapse at the end of this year. This short report provides some baseline information to inform the decision of whether the rule should be amended so that the existing standards remain, or whether they should be allowed to lapse. 1.2 Scope of this Report Although the emissions standards for both light and heavy used vehicles lapse at the end of this year, this report is concerned only with the light passenger fleet. This is because light passenger vehicles account for the vast majority of the fleet and also of annual registrations. Further, because only the rules for light used vehicles lapse at the end of the year, the rules for new light vehicles are not addressed. 1.3 Structure of this Report The remainder of this report is structured around the following key questions: Do older vehicles really emit higher rates of harmful pollutants than newer vehicles? (section 2) What restrictions do the current rules impose? (section 3) What impact have these rules had on the historic rate and composition of registrations? (section 4) What impacts, if any, have they had on the rate and composition of scrapped vehicles? (section 5) What impact would allowing the rule to lapse have on the mix of registrations and hence on the future composition of the fleet? (section 6) Have there been any wider impacts on the passenger vehicle market? (section 7) Review of Proposed Updates to Vehicle Exhaust Emissions Rules 3

2 Emission Rates for Old vs New Vehicles This section considers the local evidence on emissions rates for older and newer cars to add some context to the debate. 2.1 Results of Latest NZ Research The latest local research on emissions from light vehicles is published in NZTA research report 441, titled Are the harmful emissions from New Zealand s light duty vehicle fleet improving? (May 2011). 2 The report contains a significant amount of data collected from remote sensing sites across Auckland, and paints a very clear picture. For instance, the table below compares the recorded emissions of Japanese used petrol vehicles by the emission standard to which each vehicle was built. 3 Table 1: Emissions for Japanese Used Petrol Vehicles in 2009 by Emissions Standard The results in table 1 confirm that vehicles built to the newer standard do indeed emit much fewer pollutants. For instance, the average emission rate for carbon monoxide (CO) is 83% lower for cars built to the Japan 05 standard than those built to the much older Japan 79E standard. Similarly, Japan 05 vehicles emit 86% less hydrocarbons than Japan 79E vehicles, and 92% less nitrous oxide. While the results published for diesel vehicles are not as dramatic, they do confirm that the new emissions standard do lead to quantum reductions in the emissions of harmful pollutants, and hence that policies aimed at curbing the influx of vehicles built to these older standards is justified. 2 http://www.nzta.govt.nz/resources/research/reports/441/docs/441.pdf 3 This is a direct copy of table 5.8 on page 56 of the report. Review of Proposed Updates to Vehicle Exhaust Emissions Rules 4

3 Past and Current Emissions Rules This section briefly considers past and current exhaust emissions rules. 3.1 Timeline The following graph depicts the dates at which different exhaust emissions rules have been introduced, and also identifies the dates at which standards have been updated under the existing (2007) rule. Figure 1: Timing of Emissions Rules The first rule (the 2003 rule) was introduced in January 2004, and replaced by another rule (the 2006 rule) three years later. The 2007 rule was introduced a further 12 months down the track at the start of 2008. Since then, there have been several amendments to the Rule, including a 2010 amendment that updated the standards applying to new vehicles. Importantly, the 2007 Rule contained a timetable for introduction of a series of standards for both petrol and diesel vehicles. 3.2 Current Rules At the time of writing, used Japanese petrol and light diesel vehicles entering the fleet must be built to the Japan 05 standard, which effectively limits used registrations to a maximum age of 7 years. 4 However, this standard is due to lapse at the end of the year, prompting the current review. The standards for new light vehicles, conversely, expire at the end of 2014. These are the subject of a separate amendment, and therefore are not addressed in this report. 4 Some concessions are granted for older cars, but these account for very small proportions of registrations. In addition, some vehicles compliant with the 2005 Rule were manufactured in Japan prior to the rule being implemented. Review of Proposed Updates to Vehicle Exhaust Emissions Rules 5

% of Registrations in Each Age Band 4 Impacts on Historic Registrations This section analyses the impacts of past and current emissions rules on the rate and composition of light vehicle registrations. 4.1 Impacts on Petrol Registrations The following table and chart summarise registration data for light petrol passenger vehicles from 2000 and 2011 (inclusive). The data are grouped by emissions rule to aid comparison. Timeframe Table 2: Petrol Registrations from 2000 to 2011 (inclusive) by Emissions Rule Average Annual Registrations % of Registrations that were New % of Registrations that were Used Average Age of Registrations (yrs) Pre-2003 Rule 182,461 35% 65% 4.59 2003 Rule 202,662 37% 63% 4.79 2006 Rule 186,879 39% 61% 4.80 2007 Rule (Jap 98) 156,114 43% 57% 4.49 2007 Rule (Jap 00/02) 132,072 39% 61% 4.76 Figure 2: Age Profiles of Petrol Registrations from 2000 to 2011 by Emissions Rule 50% 45% 40% 35% 30% Pre-2003 Rule 2003 Rule 2006 Rule 2007 Rule (Jap 98) 2007 Rule (Jap 00/02) 25% 20% 15% 10% 5% 0% New 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15+ Age of Vehicle when First Registered in New Zealand (years) From 2000 to 2003 prior to the 2003 rule being introduced more than 180,000 petrol vehicles were registered each year. Just over one-third were new, and just under twothirds were used. Over the next three years, when the 2003 rule was in place, annual registrations rose and so too did the average age. Since implementation of the 2006 rule, annual registrations have steadily fallen, while the split between new and used has stayed fairly constant. However, there was an Review of Proposed Updates to Vehicle Exhaust Emissions Rules 6

Registrations (first quarter of 2012) increase in the share of new registrations under the Japan 98 standard, but this was in place for only one year. Figure 2 shows that the age distribution of petrol registrations shifted under the 2003 rule, but has not changed much since then. Whereas the age of used vehicles peaked at around 7 years prior to the 2003 rule, it now peaks at around 9 years. Unfortunately, the data above did not capture the effects of the Japan 05 standard, which was introduced only at the beginning of this year. However, the following graph shows the age distribution of petrol registrations during the first quarter of 2012. A number of these vehicles would be unlikely to meet the new standard, so were probably imported toward the end of last year. 5 As a result, registrations for the year to date are unlikely to provide a reliable indication of the future age profile associated with the updated (Japan 05) standard. 18,000 Figure 3: Petrol Registrations for the First Quarter of 2012 16,000 These registrations are unlikely to meet the updated import standards 14,000 12,000 10,000 8,000 6,000 4,000 2,000 0 2012 2011 2010 2009 2008 2007 2006 2005 2004 2003 2002 2001 2000 older Year of Manufacture Overall, the data does show a rapid decline in annual registrations, but it does not reveal any significant impacts on age profiles. While it may seem tempting to assume that the fall in registrations is due to the rules, there are several other potential forces at work. We return to this point soon. 4.2 Impacts on Diesel Registrations The following table and chart summarise the corresponding data for light diesel passenger vehicles. Again, the information is grouped by rule. 5 These discrepancies arise because the rules are applied at the time a vehicle is first certified for entry to the fleet, not at the time of actual registration, which can be several months later. Review of Proposed Updates to Vehicle Exhaust Emissions Rules 7

% of Registrations in Each Age Band Timeframe Figure 4: Analysis of Light Diesel Registrations from 2000 to 2011 by Emissions Rule Average Annual Registrations % of Registrations that are New % of Registrations that are Used Average Age of Registrations (yrs) Pre-2003 Rule 35,064 40% 60% 4.65 2003 Rule 42,551 45% 55% 4.86 2006 Rule 38,064 67% 33% 2.95 2007 Rule (Jap 02/04) 26,569 87% 13% 1.11 2007 Rule (Jap 05) 26,131 97% 3% 0.17 Figure 5: Age Profiles of Light Diesel Registrations from 2000 to 2011 by Emissions Rule 100% 90% 80% 70% 60% Pre-2003 Rule 2003 Rule 2006 Rule 2007 Rule (Jap 02/04) 2007 Rule (Jap 05) 50% 40% 30% 20% 10% 0% New 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15+ Age of Vehicle when First Registered in New Zealand (years) Like petrol registrations, diesel registrations increased following adoption of the 2003 rule, and the share of new registrations also increased. However, since then, not only has the number of registrations fallen, but so too has the proportion of used vehicle registrations. For instance, under the current rule, only 3% of light diesel registrations were used vehicles, compared to 55%% under the 2003 rule. As a result, the average age of diesel registrations has fallen to less than one year The rapid shifts in the age of diesel registrations is also illustrated in figure 4, where the proportion labelled new has increased markedly, and the rest of the curve has flattened to almost nothing. 4.3 Controlling for Other Effects As noted earlier, while it may seem natural to assume that the observed trends in registrations are driven solely by emissions rules, there are various other forces at work. For instance, between 2004 and 2008, Russia became a major importer of used Japanese vehicles, thus displacing a large pool of vehicles that may have otherwise been imported here. In addition, the Japanese earthquake of March 2011 caused major disruptions to the vehicle supply chain, while the global financial crisis has dampened demand. Fuel Review of Proposed Updates to Vehicle Exhaust Emissions Rules 8

Feb-00 Jul-00 Dec-00 May-01 Oct-01 Mar-02 Aug-02 Jan-03 Jun-03 Nov-03 Apr-04 Sep-04 Feb-05 Jul-05 Dec-05 May-06 Oct-06 Mar-07 Aug-07 Jan-08 Jun-08 Nov-08 Apr-09 Sep-09 Feb-10 Jul-10 Dec-10 May-11 Oct-11 Monthly Vehicle Imports (3-Period) Moving Average) NZ$-Yen Exchange Rate (Monthly) prices have also exhibited extreme volatility over the last 10 years or so, thus affecting vehicle demand. Beyond these factors, the other key driver of vehicle registrations is the exchange rate. For instance, between December 2000 and August 2007, the New Zealand dollar appreciated more than 120% against the Japanese Yen. However, over the following 18 months, nearly all of those the gains were wiped out. To test the strength of any relationship between the exchange rate and the rate of registrations, we first plotted the variables alongside one another. To this end, the figure below plots the impact of the NZ$/Yen exchange rate on average registrations over the following three months. Figure 6: Monthly NZ$/Yen Exchange Rate and 3-Period Moving Average of Registrations 25,000 20,000 Monthly Imports NZD-JPY Exchange Rate 120 100 80 15,000 60 10,000 40 5,000 20 0 0 Figure 6 shows that there is indeed a fairly strong correlation between the exchange rate and the rate of registrations. To further explore this effect, and to disentangle it from the effects of other factors, we ran some simple regression analyses. 6 The results are presented separately below for petrol and diesel registrations, respectively. 6 Caution must be exercised when performing regression analysis on time series data, as there is a risk that the estimated relationships are skewed by non-stationarity. Fortunately, the risks of this are greatly reduced when the regression equations themselves are firmly- grounded in theory, as is the case here. While it is possible to run various diagnostic tests to ensure that the results are robust, these are very time-consuming and hence comprehensive testing was not possible in the time available. We did, however, run some basic diagnostics, and found that the results were indeed likely to be robust. Review of Proposed Updates to Vehicle Exhaust Emissions Rules 9

4.3.1 Petrol Vehicle Regression Following is the output from running the petrol vehicle regression, which was estimated in Microsoft Excel. The data are monthly, covering the period from February 2000 to December 2011. The petrol price, exchange rate and registration variables have all been expressed using natural logarithms, so that the estimated parameters represent elasticities. This is explained further below. A variable capturing income per capita was initially included, but subsequently dropped as it was insignificant. Table 3: Regression Output for Petrol Vehicles Regression Statistics Multiple R 0.94 R Square 0.88 Adjusted R Square 0.88 Standard Error 0.07 Observations 142 ANOVA df SS MS F P-value Regression 6 4.68 0.78 169.62 2.52E-60 Residual 135 0.62 0.00 Total 141 5.30 Coefficients Std Errors t Stat P-value Lower 95% Upper 95% Intercept 7.91 0.43 18.50 0.00 7.06 8.75 PetrolPrice -0.21 0.07-2.86 0.00-0.35-0.06 Exchange Rate 0.68 0.06 11.64 0.00 0.57 0.80 2003 Rule -0.05 0.03-2.17 0.03-0.10 0.00 2006 Rule -0.17 0.02-6.83 0.00-0.22-0.12 2007 Rule (Jap 98) -0.08 0.03-2.63 0.01-0.14-0.02 2007 Rule (Jap 00/02) -0.03 0.03-0.99 0.33-0.08 0.03 The regression output shows that the petrol price, exchange rate and emissions rules variables jointly explained a lot of the variation in petrol registrations. For instance, the r-square value shows that these variables collectively explained 88% of the observed variation in monthly petrol registrations. The ANOVA table (in the middle of the output) tests the overall significance of the regression, and shows that the probability of these results being obtained purely by chance is less than one a in a trillion. Moreover, the p-values attached to the coefficients show that many are highly statistically significant in their own right. The estimated coefficients in the third table show the effects of each variable on petrol registrations. For instance, the coefficient attached to the petrol price shows that, all other things being equal, a 10% increase in the petrol price reduces petrol vehicle registrations by 2.1%. Similarly, the coefficient attached to the exchange rate means that a 10% appreciation of the New Zealand dollar leads to a 6.8% increase in petrol vehicle registrations (again, all other things being equal). Review of Proposed Updates to Vehicle Exhaust Emissions Rules 10

To reflect the way that emissions rules have been progressively introduced over time, the coefficients attached to each rule measure their incremental effect over and above all previous rules. For instance, the coefficient attached to the 2006 Rule shows its incremental effect over and above the 2003 Rule, while the coefficient attached to the 2007 rule shows its incremental impact over and above both the 2003 and 2006 Rules. Given this definition, the coefficient estimates attached to the rules suggest that, all other things being equal: The 2003 rule was associated with a 5% reduction in monthly registrations, but this effect was not statistically significant. The 2006 rule was associated with a further 17% reduction in monthly registrations, and this was of high significance. The 2007 rule (Japan 98 standard) was associated with another 8% reduction in monthly registrations, but this effect was of borderline significance, and The Japan 00/02 standard was associated with a final 3% reduction in monthly registrations, but this was not statistically significant at all. These interpretations imply that some rules have had greater impacts than others, which may well be true. However, it is important to note that these variables will also be picking up the effects of other factors temporally-aligned with the rules. For instance, the 2007 Rule (Japan 98) will also be picking up the effects of any other factors in 2008 that are not explicitly factored in the model. To the extent that these background effects reinforce or offset the effects of rules, the estimated effects of the rules may under- or over-stated. This caveat seems particularly important given that the greatest effects appear to be associated with the 2006 rule. This particular rule made no material changes to the standards for light passenger vehicles, so the effects attached to it almost certainly reflect other factors. Unfortunately, it is difficult to say any more than this without conducting a full-blown empirical analysis, which is well beyond the scope of this report. 4.3.2 Diesel Vehicle Regression Following is the output from running the light diesel vehicle regression, which was estimated using the same method as for petrol vehicles. Review of Proposed Updates to Vehicle Exhaust Emissions Rules 11

Table 4: Regression Output for Light Diesel Vehicles Regression Statistics Multiple R 0.90 R Square 0.81 Adjusted R Square 0.80 Standard Error 0.11 Observations 142 ANOVA df SS MS F P-value Regression 6 7.48 1.25 97.72 1.18E-46 Residual 135 1.72 0.01 Total 141 9.20 Coefficients Std Error t Stat P-value Lower 95% Upper 95% Intercept 4.36 0.44 9.91 0.00 3.49 5.23 DieselPrice -0.16 0.06-2.40 0.02-0.28-0.03 Exchange Rate 1.07 0.09 12.48 0.00 0.90 1.24 2003 Rule -0.07 0.04-2.02 0.05-0.15 0.00 2006 Rule -0.26 0.04-6.42 0.00-0.34-0.18 2007 Rule (02/04) -0.08 0.05-1.81 0.07-0.18 0.01 2007 Rule (05) 0.05 0.03 1.60 0.11-0.01 0.12 The regression output shows that fuel prices, the exchange rate and emissions rules also jointly explain a lot of the variation in diesel registrations. For instance, the r-square shows that these variables jointly explain 80% of the observed variation in monthly diesel vehicle registrations. Again, the ANOVA table shows that the chances of obtaining these results purely by chance are almost nil, while the p-values attached to the coefficients show that several are highly statistically significant in their own right. The estimated coefficients in the third table show that diesel registrations are potentially more sensitive to the exchange rate than petrol registrations, but less sensitive to fuel prices. For instance, the exchange rate coefficient implies that a 10% appreciation of the New Zealand dollar leads to a 10.7% increase in diesel registrations, while a 10% increase in the diesel price leads to a 1.6% reduction (all other things being equal). The interpretation of the emissions rule coefficients mirrors that for the petrol regression. Hence, the emission rules coefficients above suggest that, all other things being equal: The 2003 rule was associated with a 7% reduction in monthly registrations, but this effect was not statistically significant. The 2006 rule was associated with a further 26% reduction in monthly registrations, and this was of high significance. The 2007 rule (Japan 02/04 standard) was associated with another 8% reduction in monthly registrations, but this effect was not statistically significant, and Review of Proposed Updates to Vehicle Exhaust Emissions Rules 12

The Japan 05 standard was associated with a 3% increase in monthly registrations, but this was also not statistically significant. 4.4 Conclusion The relatively simple analysis presented in this section suggests that the rules are unlikely to have had any material impacts on either petrol or light diesel registrations, as the coefficients attached to most rules were statically insignificant. Moreover, the rule with the greatest apparent impact (the 2006 rule) actually had no bearing on the light passenger fleet, so the estimated effects attached to that almost certainly reflect a range of other factors. By far, the greatest impacts on light vehicle registrations seems to come from fuel prices and the exchange rate. Review of Proposed Updates to Vehicle Exhaust Emissions Rules 13

new 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30+ % of Deregistrations in Each Age Band 5 Impacts on Historic Scrappage Rates This section tests for any effects of the rules on the rates of vehicle scrappage. 5.1 Understanding the Links between Registrations and Scrappage While the emissions rules directly affect only vehicle registrations, they may also indirectly affect deregistrations. This is because, if the rate of registrations falls due to the rules, the rate of deregistration must also fall. Otherwise, the fleet will be smaller than it would have been otherwise. In other words, the rules may exert both direct and indirect influences. Of greatest interest here is not whether the rate of deregistrations has simply fallen, but whether the age of the scrapped vehicles has changed. Indeed, if the rules cause vehicles to stay in the fleet longer, and thus to be scrapped at an older age, there could be some adverse effects. This is explored separately for petrol and diesel vehicles below. 5.2 Impacts on Age of Petrol Deregistrations The following two charts explore any potential effects of the emissions rules on the age of petrol deregistrations. Overall, they show that there has been very little effect, if any. Figure 7: Age Profiles of Petrol Deregistrations from 2000 to 2011 by Emissions Rule 12% 10% Pre-2003 Rule 2003 Rule 2006 Rule 2007 Rule (Jap 98) 8% 2007 Rule (Jap 00/02) 6% 4% 2% 0% Age of Vehicle when Deregistered in New Zealand (years) Review of Proposed Updates to Vehicle Exhaust Emissions Rules 14

New 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30+ % of Deregistrations in Each Age Band Average Age of Vehicles when Deregistered Figure 8: Average Age of Petrol Deregistration from 2000 to 2011 by Emissions Rule 20 18 17.66 17.45 17.33 17.47 17.89 16 14 12 10 8 6 4 2 - Pre-2003 Rule 2003 Rule 2006 Rule 2007 Rule (Jap 98) 2007 Rule (Jap 00/02) Emissions Rule in Force at Time of Deregistration 5.3 Impacts on Age of Diesel Deregistrations The following two charts look for any effects of the rules on the age of diesel deregistrations. Unlike petrol registrations, the rules do appear to have affected the age of diesel deregistrations. However, the caveat raised in the previous section also applies here. Thus, while it may seem that the rules are having effects, these will also be picking up other effects not explicitly accounted for. Figure 9: Age Profiles of Light Diesel Deregistrations from 2000 to 2011 by Emissions Rule 12% 10% Pre-2003 Rule 2003 Rule 2006 Rule 2007 Rule (Jap 02/04) 8% 2007 Rule (Jap 05) 6% 4% 2% 0% Age of Vehicle when Deregistered in New Zealand (years) Review of Proposed Updates to Vehicle Exhaust Emissions Rules 15

Average Age of Deregistered Vehicles Average Age of Vehicles when Deregistered Figure 10: Average Age of Light Diesel Deregistration from 2000 to 2011 by Emissions Rule 20 18 16 15.65 16.33 16.87 17.83 14 14.06 12 10 8 6 4 2 - Pre-2003 Rule 2003 Rule 2006 Rule 2007 Rule (Jap 02/04) 2007 Rule (Jap 05) Emissions Rule in Force at Time of Deregistration To try and separate the effects of the rules from other (unaccounted) forces, we recreated figure 10 using annual data. If the rules are the key driver of increases in the age of scrapped diesels, we would expect to see step changes between one rule and another. However, if some other forces are at work, we would expect to see the average age creeping up over time irrespective of the rule in place. The results are reproduced below, where the different colours represent different emissions rules/standards. Figure 11: Average Age of Light Diesel Deregistration from 2000 to 2011 20 18 y = 0.43x + 12.89 R² = 0.99 16 14 12 10 8 6 4 2 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Review of Proposed Updates to Vehicle Exhaust Emissions Rules 16

Figure 11 seems fairly conclusive. It shows that the average age has steadily increased over time irrespective of the rules, and that there have been no obvious step changes from one rule or standard to the next. This is confirmed by the linear time trend, which has an r 2 of 0.99. On this basis, we conclude that the rules have not had any material impact on the age of scrapped vehicles, but that other factors have. Review of Proposed Updates to Vehicle Exhaust Emissions Rules 17

6 Potential Impacts of Allowing a Rule Lapse This section considers the potential impacts of allowing the rules for used light vehicles to lapse. 6.1 Discussion As noted earlier, the standards for light used vehicles expire at the end of this year. This raises the interesting question of what might occur if they were allowed to lapse. The first point to note is that the impacts of not updating the standards would depend critically on the duration of any lapse. Thus, if the standards lapsed at the start of 2013, but were reinstated in 2015 (when the standards for new vehicles will likely be updated), any effects may be minor. This is particularly true given that annual registrations are typically no more than 5% of the existing fleet. However, even if the number of older cars registered due to a lapse comprised a small proportion of the existing fleet, they could have disproportionate impacts on emissions. Indeed, older cars really do tend to punch above their weight when it comes to emissions rates. If, on the other hand, the standards were allowed to lapse permanently, the impacts could be quite far-reaching. The door would be left open for a significant influx of older, gross-polluting vehicles, the effects of which could quickly erode gains made in recent years. On these grounds, a permanent lapse of the standards would seem highly undesirable. Beyond these primary considerations, allowing the standards to lapse even temporarily would probably not be ideal because: It could undermine the Ministry s credibility as an effective regulator, and thus have unfavourable reputation effects. In addition, it could create significant resistance if/when standards are proposed for reinstatement in future. By contrast, continuing with the current standards sends a clear signal that they are here to stay. As a result, efforts to move to tighter standards in future may be more straightforward To estimate the impacts of allowing the standards to lapse, even temporarily, we created a simple fleet-turnover model to test how the future fleets might look with and without a lapse. While such an exercise is not particularly difficult, the results are highly sensitive to assumptions. For instance, the results depend critically on both the assumed future profiles of registrations and scrappage. To complicate things further, changes to the fleet resulting from a lapse tell only part of the story. This is because the size and composition of the fleet is only one of several factors that determine the adverse effects of harmful emissions. Consider the following diagram. It shows that the effects of emissions depend only partly on the fleet. Other important factors include how and where the fleet is utilised. For instance, an increase in harmful emissions in remote areas is likely to have fewer adverse health impacts than Review of Proposed Updates to Vehicle Exhaust Emissions Rules 18

increased in heavily urbanised area. Similarly, an increase in the number of older cars may be less of a problem if they travel infrequently. Figure 12: Factors Influencing Emissions Impacts The take-away message is that the harmful effects of vehicle emissions are determined by a complex web of interactions, only some of which depend on the composition of the fleet. Consequently, there is limited value in trying to model the incremental changes in fleet composition that might accompany a lapse of the standards, temporary of otherwise. For these reasons, it seemed prudent to avoid any quantitative analysis, and instead rely on the brief qualitative analyses set out above. Review of Proposed Updates to Vehicle Exhaust Emissions Rules 19

Change of Ownership Transactions Change of Ownership Transactions 7 Review of Wider Effects This section briefly considers two potential wider effects of the rules. 7.1 Impacts on Market Activity An interesting consideration is the extent to which the rules may have affected general activity in the passenger vehicle market. To gauge this, we used data on changes in ownership transactions from 2005 to 2011 to look for any patterns. The results for the sale and purchase of cars are shown below. Figure 13: Trends in Change of Ownership Transactions for Cars 1,000,000 900,000 800,000 2006 Rule Adopted 2007 Rule Adopted 700,000 600,000 500,000 400,000 300,000 200,000 100,000 0 2005 2006 2007 2008 2009 2010 2011 Figure 14: Trends in Change of Ownership Transactions by Type for Cars 500,000 450,000 400,000 350,000 2006 Rule Active 2007 Rule Active 300,000 250,000 200,000 150,000 100,000 50,000 0 2005 2006 2007 2008 2009 2010 2011 Public to public 440,000 456,000 475,000 462,000 454,000 456,000 450,000 Public to trader 159,000 149,000 138,000 125,000 125,000 132,000 135,000 Trader to public 231,000 222,000 213,000 187,000 177,000 176,000 178,000 Public to public Public to trader Trader to public Review of Proposed Updates to Vehicle Exhaust Emissions Rules 20

Figure 13 shows that there was a reduction in the total number of ownership transactions following implementation of the 2007 rule, however it is impossible to say whether the rule played any real part in this. Figure 14 provides some additional context. It shows that there was a noticeable increase in private sales between 2005 and 2007, but a steady reduction thereafter. Interestingly, the number of sales by car yards fell each year from 2005 to 2009, but has remained fairly constant since then. Trade-ins have exhibited a similar pattern, albeit with a slight upswing in recent times. Overall, the results are inconclusive. There is no solid evidence that the rules have impacted the general level of activity in the vehicle market. 7.2 Impacts on the Value of the Existing Fleet The other issue is the potential impacts of the rules on the value of the existing fleet. Specifically, some commentators have claimed that the rules have increased the price of the existing fleet by restricting the rate of new supply. While data and time limitations have precluded a detailed analysis of this, a brief discussion seems warranted. First, while it is true that a restriction in the rate of supply is likely to increase the value of the existing fleet, all other things being equal, there are myriad other supply and demand forces at work. It would therefore be difficult to prove that any observed price effects are caused solely, or principally, by the requirements of the rules without first controlling for all other possible effects a difficult task indeed. Second, even if it could be shown that the rules did have inflationary impacts on the existing fleet, the effects would largely cancel out. This is because the higher price paid by consumers for their next car would be offset (at least partially) by the higher prices they would receive from selling their existing cars. Indeed, only first-time buyers would be affected in net terms. This raises the interesting question of how prices have actually changed over the last few years (irrespective of whether we can prove any links to the rules). To this end, the following graph shows the average sales price of used vehicles on Trademe in 2006 and 2011 by vehicle age. All prices are expressed in real terms using 2006 dollars. Prices for newer vehicles are on the left, while those for older vehicles are on the right. The graph shows that there has been a general increase in vehicle prices, particularly those aged 7 to 10 years. However, the price differentials for older vehicles (displayed towards the right of the graph) are quite small. In fact, the prices for the oldest vehicles (labelled 26 years old) were lower in 2011 than in 2006 in real terms. Now, since most first time buyers are likely to be purchasing vehicles from the older end of the fleet, and because prices for these older cars are about the same now as they were in 2006, any effects of the rules on first time buyers are likely to be minor. This is especially true given that first time buyers comprise a very small proportion of the overall market, and that many receive financial assistance from parents/guardians. Thus, any effects on the value of the existing fleet are likely to be minor overall. Review of Proposed Updates to Vehicle Exhaust Emissions Rules 21

Average Sales Price on TradeMe Figure 15: Average Sales Price on Trade Me in 2006 and 2011 (in 2006 $) $14,000 $12,000 2006 2011 $10,000 $8,000 $6,000 $4,000 $2,000 $0 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 Age of Vehicle at Time of Sale (Years) Review of Proposed Updates to Vehicle Exhaust Emissions Rules 22

Appendix: Schedules from the 2007 Rule The following excerpts from the Land Transport Rule: Vehicle Exhaust Emissions 2007 show that no requirements are currently scheduled for used vehicles from 2013 and beyond. Review of Proposed Updates to Vehicle Exhaust Emissions Rules 23