Labor Senators Dissenting Report

Labor Senators Dissenting Report

History of GROCERYchoice

1.1        Prior to the last election the Government acknowledged the concerns of Australian consumers in regards to what have been referred to as "kitchen table" issues such as grocery and petrol prices.  On 11 July 2007, the Prime Minister gave the following commitment:

First, if Labor forms the next government of Australia, the ACCC will be directed to publish a periodic survey of grocery prices at supermarkets for a typical shopping basket, including family staples like biscuits, bread and baked beans.  The ACCC will establish a dedicated website to publish this pricing snapshot – this very act will serve to increase transparency in the market place and in so doing exert greater competitive pressure on the retail market. Surveys will be conducted in all states, in both metropolitan and regional areas, and will provide the ACCC with the pricing information it needs to identify whether there are indications of breaches of the Trade Practices Act that require further investigation.[1]

1.2        Contrary to claims by the Coalition, the Rudd Government did not promise to reduce grocery or petrol prices but did commit to a reform agenda to remove barriers to competition, increase the powers of the regulator, and improve price transparency for consumers.  The Government has proceeded to do this by introducing a number of legislative reforms in the area of consumer law, which are outlined at the conclusion of the report.

1.3        The GROCERYchoice website was established with the aim of providing a site where consumers could access a comparison of timely, reliable prices for local stores all around Australia.  The objective was to enable consumers to easily locate the cheapest store in their area for a particular basket of goods. 

1.4        The establishment of a website such as GROCERYchoice has not been undertaken previously by any Australian Government. In the view of Labor Senators, it needs to be recognised that the establishment of GROCERYchoice was a significant and complex challenge that the former government did not have the political will to undertake.

1.5        Furthermore, whilst it is very disappointing that the website was not able to meet the Government's objective of providing timely, reliable and local pricing, Labor Senators believe that testing of the GROCERYchoice website was warranted given the importance of price transparency for consumers.

1.6        Early in 2008, the Government asked the ACCC to undertake a monthly survey of grocery prices for a typical shopping basket of goods across Australia and to establish a dedicated website on grocery prices. The ACCC ran the website for a period of six months during which time it received pricing data on 500 items from 600 supermarket outlets in sixty one regions every month and posted a summary of this information on its website.

1.7        To prevent any manipulation by supermarkets, the list of 500 products remained confidential to the ACCC.  While there were 500 products included in the price survey each month, a smaller sample of the products were used to calculate the monthly basket prices published.  Each month, changes to the product sample were made so that the products contributing to the basket prices were gradually rotated each month.

1.8        From its inception it was acknowledged that the ACCC website would have some limitations.  The Chairman of the ACCC, Mr Graeme Samuel, commented that GROCERYchoice would provide new information to consumers but would not publish the weekly specials or list grocery prices for individual supermarkets.[2]

1.9        The Government recognised that the absence of weekly specials on the site and the update of data monthly did not go far enough in providing consumers with the timely, accurate and local pricing information they desired.

1.10      However, claims that the site was a waste of money fail to recognise that the ACCC successfully ran and operated the GROCERYchoice website for six months and, in that time, met all of the requirements that the Government had committed to deliver and offered some improved price transparency for consumers.

1.11      The ACCC was always limited in relation to the information it could present on the website. Adding new features like ‘specials' would have compromised the regulator, whose responsibility it is to investigate false or misleading representations in advertising.  Therefore, in December 2008, the Government entered into a contract with Choice to deliver the GROCERYchoice website and expand the information it would provide.

1.12      The contract required Choice to deliver on the following four outcomes:

a) Publication of basket prices for individual supermarkets locations, with the contents of the baskets transparent to users;

b) Publication of basket prices for all leading supermarkets chains or groups (e.g. Coles, Woolworths, IGA, Franklins, Supabarn, Foodworks, ALDI, etc) and other smaller independent operators;

c) Publication of a ‘staples’ basket or something similar, that allows for a suitable comparison of prices between the larger supermarket chains, smaller independent supermarkets and ALDI;

d) Publication of grocery prices on a no less than weekly basis with the ‘date updated’ clearly indicated on the Grocery Choice website.  All data that is out of date or has not been updated in over a week must be removed from the GROCERYchoice website.[3]

1.13      The contract was terminated on 26 June of this year on the grounds that Choice was unable to deliver on all of the outcomes outlined in the contract.  In particular the requirement of providing basket prices for individual supermarkets would not have been met.

1.14      The Government has been upfront about the reasons why the contract with Choice was terminated.  Consumers who relied on the website might feel misled if its information was not accurate.  In a media release on the 26 June 2009, the Hon Dr. Craig Emerson MP said that:

The Government remains of the view that consumers are better placed to make informed choices when they are able to gain access to prices conveniently and make comparisons among supermarkets. However, the Grocery Choice proposal as originally envisaged would not be able to generate reliable, timely data as a basis for consumers to make meaningful comparisons in their local neighbourhoods.[4]

1.15      At the time of the launch, the Government made it clear that through GROCERYchoice it wanted to give a guide to consumers as to the cheapest supermarkets in their region.[5]  Unfortunately, the Choice site would not have delivered on this commitment and as a result the Minister determined not to continue with the Choice website.

1.16      Furthermore, Labor Senators note that the contract with Choice allowed for the cancellation of the contract at any time by Treasury.  Clause 17.1.1 of the contract states: 'In addition to any other rights it has under this contract, Treasury may by notice, at any time and in its absolute discretion, terminate this contract or reduce the scope of this contract.'[6]

Criticisms of GROCERYchoice


1.17      The total cost for the GROCERYchoice measure was $7.7 million dollars with an estimated contingent liability of $700,000.  Labor Senators note that this did deliver a website which received at its lowest usage 260,000 hits or 61,000 page views in a four week period.  Consumers did receive improved price transparency as a result of the site and there is some evidence to suggest that some downward pressure on grocery prices did occur as a result of the website.

1.18      Both Professor Allan Fels, the former ACCC Chairman, and Professor David Cousins commented that 'despite its severe limitations' the grocery price monitoring activity undertaken by the ACCC:

...seemed to have some beneficial impact for consumers, as the major suppliers seemed to respond to the favourable publicity for ALDI's low prices by also attempting to be the lowest price setters in a region.[7]  

Labor Senators reject the assertion that the website did not offer some value during its operation and believe that even a small increase in pricing transparency and any downward pressure on grocery prices is beneficial to consumers.

1.19      Labor Senators also note that to put the expenditure on GROCERYchoice into some perspective, the ACCC's total net resourcing as proposed by the May 2009 Budget for 2009–10 is more than 140 million dollars.[8]

Accessibility of Real Time and Local Pricing Data

1.20      Choice’s proposal was for prices of at least 5,000 items to be provided on a weekly basis. Since there are well over 4,000 supermarkets in Australia, that could involve more than 20 million prices being checked and provided every week.

1.21      The committee heard evidence from the major retailers and their industry representative, the Australian National Retailers Association (ANRA), that the data being requested by Choice was not able to be provided in the form required by Choice.  ANRA noted that the retailers had significant concerns regarding price accuracy.  They used the example of soft drink to explain what data is recorded by stores locally.

On the issue of price accuracy, say you have 100 units of a particular soft drink to be sold and there is—as there has been today—an unseasonable cold snap. Say four of those bottles of soft drink are stolen, because that does happen, and four of them might have been damaged, as also happens. A proportion of the rest are discounted heavily because of the local conditions—that is, the weather—to clear the stock. The data then supplied to head office on a regular basis reflects the number of units sold against the total revenue figure. It is not necessarily represented in an immediate sense on a price-per-item basis. Aldi are actually able to do slightly different things because they do have uniform national prices for products across their 500 lines, compared to the 20,000 or 30,000 lines that you would find in a large supermarket.  Any average price derived by this process would be inaccurate at best. What we were asked to do by Choice was to provide a price that nobody actually pays, which might be different to our advertised price and would potentially be misleading to consumers. This is the nub of the problem with Choice’s model. It was flawed to start with. It was not based on the commercial reality of pricing operations at store level, or the data collection ability of the major retailers. What Choice asked our members to send to them would not have been accurate according to their specifications. [9]

1.22      During the hearings, the committee heard concerns that the major supermarkets through their representative may have worked together to ensure that the data was not made available to Choice.  Senator Joyce went further in a comment to the ACCC:

They are 70 or 80 per cent of the retail market and they are working together in a coordinated form to basically sink the GROCERYchoice website.[10]

1.23      Labor Senators don't believe that the committee received any evidence to suggest that the major supermarkets deliberately colluded to prevent the supply of pricing data.  Virtually all of the submitters who voiced concerns regarding the ability of the majors to supply the data assumed that the retailers were able to supply the data and were choosing not to do so.

1.24      Woolworths refuted this when appearing before the committee:

It comes down to this: the Choice model assumes that Woolworths—and I cannot comment for any other retailer—has a central database on a computer that sits in our head office and can tell us exactly the price of a good as it goes through the register and sells in real time. We had to explain to Choice in December that we do not have that.[11]

1.25      Coles commented that:

We do not monitor real-time price data across all of our supermarkets. Supermarket managers have the flexibility to price match for local competition, and therefore prices can vary daily across our supermarket network. So we do not have that data.[12]

1.26      In response to Choice's claims that ANRA's involvement in the GROCERYchoice negotiations was anti-competitive Coles responded:

I think that is completely false. The motivation for involving ANRA towards the end of the discussions between Choice and the supermarkets was to expedite and facilitate discussions around very complex and difficult issues where we could get the insights of other members about how best to resolve the issues. It is extremely disingenuous to say that that was anticompetitive. In fact, the intention was to be cooperative and to try and meet a very looming deadline.[13]

1.27      ANRA themselves responded that:

Far from the spoiler role suggested by Choice, we were attempting to bring some focus to the range of issues that, just two months out from the launch of the site, Choice had still not dealt with or given any real to clarity to the retailers on.[14]

1.28      Labor Senators acknowledge the work undertaken by Choice to prepare the GROCERYchoice website. However, we not believe there is sufficient evidence to show that ANRA deliberately acted on behalf of the major retailers to prevent the supply of requested pricing data.  Labor Senators take the view that Choice designed the website with an entirely consumer centric focus and as a result it began to incorporate a wide range of data not previously considered.  Woolworths noted that, as the project developed, the information being sought by Choice became more complex:

What then happened was that as the project progressed Choice started adding more and more complication to it. They started asking us for people who could provide data feeds on the type of trolleys that each store had, so across 800 stores we were required to tell Choice what types of trolleys we had. They asked us to identify stores and their public transport options. They asked us to identify whether we had on-grade car parking. They then said as a measure of course they were going have consumers do their own surveys of the prices published and they would then come back to us and ask us to provide explanations as to any discrepancies between what was published and what they saw the website. So we then ended up with a headcount of at least half a dozen people who were going to have to support the project.[15]

1.29      Choice also made the incorrect assumption that real time pricing data was uploaded from the point of sale in local supermarkets to the retailers head office.  It is disappointing that these issues were not addressed earlier in the negotiations, as Labor Senators are of the view that had they been overcome then the Choice website would have contributed to improved price transparency and provided a useful guide for consumers.

Independent Retailers

1.30      The committee heard criticism from smaller independent retailers that the GROCERYchoice website was prejudicial and did not represent them fairly.  They argued that issues such as lack of competition in remote areas and the high costs of delivery (particularly for fresh produce) were not adequately accounted for and therefore resulted in 'negative press' about independents once the survey results were published.[16] 

1.31      They also complained that the ACCC included stores under 1000 square metres in the survey although this was not the methodology the ACCC had outlined in their approach to the surveying.

1.32      Responding to the claim that the website had favoured the major retailers over the independents, the ACCC noted that the grocery inquiry report had highlighted the difficulties that independents have in competing with Coles and Woolworths – largely because of the prices they are charged by their wholesaler Metcash.  They also pointed out that, during the six month operation of GROCERYchoice, approximately 360 regional readings were taken and in only approximately eight of these cases did the independent stores come out cheapest.  The ACCC noted that this supported the results of the grocery inquiry, and indeed the arguments put by the independent retailers themselves.

1.33      The ACCC also noted, in regards to stores being included with floor space under 1000 square metres, that 'they were the exception rather than the rule'[17] and that those stores 'identified by industry as being smaller than desirable [were] removed from the survey list'.[18]

1.34      Whilst acknowledging the concerns put by independent retailers, Labor Senators do not believe that the GROCERYchoice website drastically misrepresented independent grocery stores and their competitiveness and supports the ACCC's findings in the grocery inquiry that smaller independents often face higher supply costs than the major chains.

Tender Process and Planning

1.35      Coalition Senators are critical of the timeframes the ACCC were given by the Government to establish the GROCERYchoice website.  However, the ACCC in evidence to the committee noted that they went for the cheapest tender that would also deliver accurate data and that a normal procurement process was adopted meaning that a best value for money approach was taken.[19] 

1.36      Furthermore, the ACCC undertook a scoping study in order to assess the best way of implementing the Government's election commitment in regards to GROCERYchoice, and the ACCC told the committee that "the site that went up in early August was fairly close to what we put to the government as being the preferable way of going."[20]

1.37      Coalition Senators also criticised the tender process itself, due to the submission of Informed Sources, which at its core criticised the ACCC evaluation panel on the basis of the company's previous opposition to the Government's Fuelwatch scheme.

Whilst we are not suggesting there was any inappropriate direction from the Minister responsible, the Chairman, the Chief Executive or the Management of the ACCC, there was clearly a degree of “negative interpretation” or group think operating within the evaluation panel.[21]

1.38      Coalition Senators claim that the ACCC spent $2.7 million more than necessary in the tendering process, but conveniently ignore evidence provided by the ACCC regarding their concern that Informed Sources may not have been able to provide accurate data in the timeframes required.  Contrary to the views expressed by Coalition Senators, we believe that the ACCC showed appropriate due diligence by ensuring that the tender was not just assessed on cost but importantly on the certainty of delivery of timely and accurate data.

1.39      Labor Senators acknowledge the statement by Informed Sources that the evaluation panel assessment was incorrect – however, we fail to see any evidence of 'negative interpretation' or bias against the company, particularly given the ACCC's renegotiation of another contract with Informed Sources for the supply of fuel pricing data that occurred concurrently.

Integrity of Data Collection by the ACCC

1.40      During the inquiry, concerns were raised regarding the integrity of the data collection undertaken by the ACCC for its GROCERYchoice website.  This criticism was twofold: firstly, that the successful tendered Retail Facts also undertook data collection for Woolworths and secondly, that the ACCC did not undertake an on site spot check audit of any of the data during the operation of the website.

1.41      It is important to note that both Informed Sources and Retail Facts collected pricing data for major retailers.  The ACCC pointed out that whilst Retail Facts collected data for Woolworths, Informed Sources collected data for Franklins.  Both are retailers that would also be covered by the GROCERYchoice website. 

1.42      The criticism was that there may be some conflict of interest in Retail Facts gathering price data for both Woolworths and GROCERYchoice and that the data may somehow cross over, enabling Woolworths to know what items made up the basket and price them lower accordingly to 'game' the system.  There were also concerns that Retail Facts would possibly just do the one price check and charge both customers for simply doing one lot of work.

1.43      These concerns were addressed by Retail Facts during the public hearing.  Firstly, Retail Facts noted that the data collecting for each contract is always done separately – that data is never collected for two customers at the same time by the same person.

A point that supports this is that throughout the 11 months that Retail*Facts conducted the price collections for the ACCC we visited some 6,255 stores across Australia and we collected details on approximately 2,400,000 products, which totalled, for the record, nearly 12 million at the points. It also should be noted that during the price collection process, throughout this massive undertaking, there was not one instance of dual collecting presented to us or to anybody that I am aware of. We are very proud of and would put on record the job we have done with the ACCC.[22]

1.44      Secondly, Retail Facts vehemently rejected any notion that data would cross over between clients.  They argued that the highest levels of confidentiality were maintained and that this negated any potential conflicts. Furthermore, they noted that not one example had been provided of a Retail Facts employee not acting with integrity and meeting their commitments under confidentiality agreements:

The key points to support this are—and some of these have been mentioned by the ACCC—that every employee in the Retail*Facts team was committed to the undertakings required by the ACCC and signed confidentiality agreements to that extent. In addition, we developed proprietary systems to provide specific product information to the price collectors and had that information available on a limited time. In most instances it was no more than 48 hours when that information was then taken away from their identification. Also to ensure confidentiality, specific and separately designed databases were used for each customer.[23]

1.45      Labor Senators also note the ACCC's point that it is not realistic to expect that a data collection agency will never have other clients who may be viewed as being in competition for the information and that this is why confidentiality agreements are regularly used in these circumstances:

They are under contractual obligations to us. We encounter this quite often. It is not realistic to think, ‘If someone else is using somebody to do something then we will not.’ That is why we have safeguards and obligations in our contracts about confidentiality and so forth. It would almost be inevitable that some of the same data was being collected. Given we are collecting across 500 grocery items, there would be some commonality but only some in the sense that Woolworths or whoever else would be interested in some of the same items and some different items.[24]

1.46      Another contracted company, FreshLogic, when asked about the need for additional mechanisms to ensure there is no transfer of information responded by saying that:

No, I do not believe we need that. We work with confidentiality agreements, particularly when we get down to advising people on business strategy. There may then be a competitive variable, and we would deal with that through a confidentiality agreement.[25]

1.47      Labor Senators do not believe that any evidence was proffered to the committee to suggest that the data collection process for GROCERYchoice was in any way jeopardised as a result of a breach of confidentiality.  We note that the most important factor is retaining the confidentiality of the items being price checked.  Given the rotation of dates, stores and the basket itself, combined with the systems implemented by Retail Facts to ensure confidentiality, Labor Senators believe the opportunity for 'gaming' would be very small. 

1.48      Retail Facts confirmed to the committee that their contract with the ACCC allowed for access to both infield audits and systems audits at any time and under any circumstances. The ACCC explained that they conducted a crosscheck of the data they received using tools audited by Frontier Economics.  The team ran through a couple of thousand items of data per month looking for any indications that the data may not be accurate.  During the time GROCERYchoice was operational they found no indication that the data was inaccurate and, therefore, did not undertake a field or systems audit.

1.49      Labor Senators are of the view that the internal data cross check used by Retail Facts combined with the further check undertaken by the ACCC would have been adequate over the time period the website operated.  However, we do find it surprising that given the contract with Retail Facts allowed both infield and systems audits 'at any time and under any circumstances'  and the fact that the contract was a reasonably significant one, the ACCC did not avail themselves of the opportunity at some point during the operation of GROCERYchoice to conduct such an audit.  Labor Senators believe this should have occurred and would have ensured full public confidence that no conflict of interest or breach of confidentiality had occurred.

Government Measures So Far to Improve Competition in the Sector

1.50      In addition to GROCERYchoice the Government has undertaken a number of reforms to improve competition in the grocery retailing sector.  Competition is by far the most effective means of exerting downward pressure on grocery prices.  The ACCC grocery inquiry found that effective barriers to entry have stifled competition and consumers have suffered as a result.

1.51      On the 18th of September of this year, the Minister for Competition Policy and Consumer Affairs, the Hon. Dr Craig Emerson MP, released a policy statement on improving competition in the retail sector.[26]

1.52      The policy statement details a number of ways in which the Government is removing barriers to entry to increase the number of competitors in the grocery sector, as well as further reforms to improve price transparency and to provide the regulator with greater powers to prosecute breaches of the Trade Practices Act 1974.

Foreign Investment Rules

1.53      Until early in 2009, it was a condition of approval that foreign investors commenced construction on vacant land acquisitions within 12 months of receiving purchase approval.  The Government extended this timeframe for development to five years recognising that 12 months was insufficient to enable a development to commence and that large companies make forward investment decisions well beyond a 12 month timeframe.

1.54      The policy benefits of this change have resulted in the planned expansion of ALDI from more than 200 stores to 700 stores and facilitated the growth of Costco from one store in Melbourne to plans for stores in New South Wales, Queensland, South Australia, the Australian Capital Territory and other regional centres.

Restrictive Lease Provisions

1.55      Restrictive lease provisions between major supermarket chains and shopping centre owners prevent centre managers from leasing space to competing supermarkets.  The ACCC has reached an agreement with Coles and Woolworths to end existing restrictive provisions in supermarket leases with shopping centres.  Restrictive lease provisions will end immediately for stores that have been operating for more than five years, and for newer stores they will phase out and end no later than five years from the stores opening date.

1.56      This new agreement will mean that out of 750 restrictive leases more than 80 per cent will cease immediately and the remaining 20 per cent will be gone within five years.  This will open up shopping centre space for competitors such as ALDI, Franklins, FoodWorks and IGA, and will directly benefit consumers by increasing local competition.

Planning Laws

1.57      Planning laws affecting the location and use of land for specific activities are implemented by the zoning of land.  By restricting the availability of retail space, state, territory and local government planning laws can hinder supermarkets from being established in particular areas.

1.58      This can be a particular problem for independent supermarkets since most states and territories have adopted 'centres' policies to concentrate retail activities in one location.  Although planning laws and zoning restrictions are necessary for traffic management, public amenity and environmental protection, they can also be used to stifle competition.  This is most apparent when existing businesses object to applications for entry by competitors.

1.59      The Government has referred the anti-competitive impacts of zoning and planning laws to the Business Regulation and Competition Working Group established by the Council of Australian Governments (COAG).  The working group will present their recommendations to COAG this year to ensure planning laws do not impede competition in grocery retailing.

Unit Pricing

1.60      The Government introduced mandatory unit pricing in Australia for larger supermarket stores—all major stores have until December 2009 to complete the roll out of unit pricing.  Unit pricing assists consumers by enabling them to compare prices readily between different sizes of the same product, different branded products and also between stores by stating the price per unit measurement (e.g. weight) so consumers can assess value by quantity.  This measure has proven to be a useful tool for consumers and enhances price transparency between products in supermarkets.

Stronger Predatory Pricing Rules

1.61      In response to the ACCC grocery inquiry the Government passed legislation to strengthen predatory pricing rules so that the major supermarkets can't easily drive out rivals by deliberately sustaining non-profitable pricing for an extended period of time to drive smaller rivals out of the market, only to then increase prices later.

Emergence of New Technology

1.62      During the inquiry, there was much discussion about the role that new technology may play into the future in improving price transparency for consumers and improving their access to information.

1.63      With mobile phone technology advancing at a rapid pace and new social networking websites such as Facebook allowing consumers to share information with thousands of other consumers in real time – arrangements whereby consumers themselves may establish, update and share information about bargains they have found or specials in their local area may become increasingly common.

1.64      Increasingly consumers are also purchasing more of their groceries over the internet.  If this trend continues to increase in popularity, we may see the emergence of a whole spectrum of "internet retailers" into the market which will benefit consumers through increased price competition. 

Industry Website

1.65      In his media release announcing the decision not to proceed with the Choice website, the Hon. Dr Craig Emerson MP said:

I will hold discussions with supermarket chains about the possibility of an industry website capable of providing convenient grocery price data that could be audited by a government-appointed auditor.[27]

1.66      During the inquiry, reference was made to ongoing conversations between ANRA and the Government regarding an industry operated and run website to provide pricing information to consumers.  Woolworths told the committee that:

The government is taking a keen interest in where retailers are going with online.  As I mentioned to Senator Pratt, we are doing a lot of work on it. I hope you will all be really thrilled with the product that Woolworths will be able to deliver to the market fairly soon. My view and the Woolworths view are that, if you look around the world, the advances that are happening at the moment in the online space are happening fairly quickly. We are starting to learn that. We are starting to look at the technology we can get and to get those brought into the Australian marketplace. I think you will see that online participation area slowly increase.

Senator BARNETT—With that Woolworths initiative, has the government approached Woolworths or other major retailers, to your knowledge, to establish an industry based website?

Mr Hall—They have been having those discussions with ANRA, as I understand it. As a member of ANRA we have been participating in those discussions at our regular ANRA meetings. But of course it is a point of competitive advantage that we are not going to talk about what we are doing as a company in this space until we are ready to release it to our customers. [28]

Labor Senators believe that this firmly supports the view that the GROCERYchoice website has been worthwhile and that more will be undertaken in this area over time to improve price transparency for consumers. 


1.67      Labor Senators are disappointed that the proposal for Choice to operate the GROCERYchoice site did not come to fruition.  However, we support the Minister's decision given the site would not have delivered accurate, timely and local pricing data enabling consumers to determine the cheapest store in their local area for an average basket of goods.

1.68      The Choice site was ambitious in trying to deliver the best outcome for consumers and Labor Senators believe that, if the data delivery issue could have been overcome earlier, the site would have been of great benefit.

1.69      However we also believe that the work done by Choice will not be wasted, and that discussions taking place with the major retailers to establish a similar site, whether such a site is taken up by the private sector or revisited by the Government, will benefit substantially from the lessons learned by the GROCERYchoice experience so far.


Senator Annette Hurley
Deputy Chair
Senator Louise Pratt

Navigation: Previous Page | Contents | Next Page