Retail electricity prices in New Zealand: recent trends and the relationship to market shares

Abstract

This paper constructs a monthly price series for electricity retailers in New Zealand, broken down by region and retailer, from January 2018 until June 2023. Since 2014, there has been no such price series for New Zealand that is disaggregated by retailer. We find that, over the time period of our dataset, while average prices across the largest five retailers are consistently higher than that of the smaller retailers in our dataset, the gap has narrowed substantially. The narrowing of this price gap is associated with the following trends in market shares. While, from January 2018 to September 2021, the smaller retailers, taken as a group, made considerable gains in market share at the expense of the larger retailers, the growth in the market share of the smaller retailers in our dataset stagnated in the remainder of the time period.

Keywords:

• Retail electricity market
• prices
• market share
• New Zealand
• competition

Acknowledgments

We would like to thank the following for their comments and assistance: Ezra Barson-Mclean; various employees at Powerswitch, especially Paul Fuge and Stephen Burrough; various employees at the New Zealand Electricity Authority, especially Phil Bishop.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 New Zealand Government (2019).

2 For a description of the recommendations that have been implemented, see Ministry of Business, Innovation and Employment (2023b).

3 Throughout the paper, when market shares and prices for a month are cited, they are to be interpreted as the value at the beginning of the month. Further, market shares are measured by the number of connections, or ICPs (installation control points). The data on ICPs are taken from the ‘Market share trends’ and ‘Market share breakdown’ tables of Electricity Authority (2023). The ‘Market share trends’ table is available at https://www.emi.ea.govt.nz/Retail/Reports/R_MST_C?_si=v—3 and the ‘Market share breakdown’ table at https://www.emi.ea.govt.nz/Retail/Reports/HR5D1V?_si=v—3 (accessed 2 September 2023). For all market shares presented below, they are calculated for residential connections. The market shares calculated above, however, are not restricted to the residential segment, because data specific to this segment is not available as early as 2010.

4 These statistics on switching exclude ‘switches’ arising when the retailer for an ICP changes as a result of a move; in such a case, while there has been a ‘switch’ in the retailer associated with the ICP, the household who moved in and the household who moved out may both have retained their previous retailers. In particular, the switching data we use are referred to as ‘trader switches’ in our data source, which obtained from the ‘Switching trends’ table of Electricity Authority (2023). This table is available at https://www.emi.ea.govt.nz/Retail/Reports/R_SwT_C?_si=tg—consumer-switching,v—3 (accessed 2 September 2023).

5 See Ministry of Business, Innovation and Employment (2023a).

6 New Zealand Government (2018, p. 38).

7 The dataset begins in January 2018 because the addition of earlier months would have created considerably greater challenges for ensuring consistency over time in our price series. The dataset ends in June 2023, because we began our project soon afterwards.

8 In a standard Bertrand model with differentiated products, if one firm lowers its price, the optimal responses of other firms are to lower their prices. See Belleflamme and Peitz (2015, p. 50).

9 For econometric models of consumer choice in retail electricity markets with switching costs, see Daglish (2016), Dressler and Weiergraeber (2023) and Hortacsu et al. (2017).

10 As specified in Powerswitch's raw data, plans sometimes offer one or more discounts on the total monthly bill. The discount is denoted by D in Equations (1(1) $P_{\mathrm{AI}}=(1-D)(\mathrm{Vx}+\mathrm{Lx}+F)$(1) ) and (2(2) $P_{\mathrm{CU}}=(1-D)(V_U{x}_U+V_C{x}_C+\mathrm{Lx}+F)$(2) ). Examples of such discounts include discounts for ‘prompt payment’ and also discounts for ‘electronic payment’. Our calculations of monthly prices are made on the assumption that the consumer takes advantage of such discounts. If, in a given month, a plan offers multiple discounts, we sum them together to calculate D. If, in a given month, a plan offers no such discounts, then D = 0.

11 For Electric Kiwi, there is no decomposition into controlled and uncontrolled consumption; even when meters are configured for CU plans, Electric Kiwi charge the same price for controlled and uncontrolled usage. Also, Electric Kiwi do not present the Electricity Authority levy as a separate tariff component; it is covered by its variable price.

12 The Big Five price series is a simple average of the five component series from Figure 1; the Small Five prices are also a simple average. But similar trends are obtained if, instead, we calculate Big Five and Small Five average prices as weighted by the market shares of the component retailers.

13 We only include the 319 plans for which the retailer was operating in the relevant region at the beginning of our time period.

Additional information

Funding

This work was supported by Otago Business School, University of Otago [Commerce Research Grant - Round 1, 2022].