[Previous] [Contents] [Next]

The Economic Impacts of Privatization

The privatization of liquor retailing in Alberta has had quantifiable impacts on liquor store locations, product selection, prices, wages, employment, and government revenues. These impacts will be examined in detail in this section.

Liquor store locations

Table 3 contains some data summarizing the expansion in the number of liquor stores in Alberta. The first two lines of the table show that there were 205 ALCB stores and 53 beer and wine stores in Alberta in August 1993.¹⁶ By December 1, 1995, there were 115 licensed, private liquor stores in Calgary, 100 in Edmonton, and 390 in the rest of Alberta. The total number of private liquor stores by December 1, 1995—605—is a 134 percent increase over the combined number of ALCB stores, wine stores, and beer stores as of August 1993.

Many of the ALCB stores were purchased and converted to private liquor stores. In Calgary, all 24 of the ALCB stores were converted to private stores, while in Edmonton 20 out of 23 were converted, and in the rest of Alberta 71 of 158 were converted. The high conversion rates in Calgary and Edmonton are partly explained by the fact that these cities were slow to issue new business licenses in the months just following privatization. In order to set up in the liquor business prior to the 1993 Christmas season, an existing ALCB store had to be purchased and converted to a private liquor store. The relatively small number of ALCB stores converted to private stores in the rest of Alberta can be explained by the fact that in many small communities, the ALCB store was one of the highest priced retail properties in the community. Many ALCB stores were also regarded as too large given the increased competition in liquor retailing. While it might be possible for a private operator to cover the costs of owning and operating a former ALCB store if there is only one liquor store in town, it might be impossible to cover these costs if there are four or five. Many of the ALCB stores in the rest of Alberta apparently were more highly valued in alternative (non-liquor store) uses.

Table 3: Store and Sales Comparisons: Pre-and Post-Privatization
	Calgary	Edmonton	Rest of Alberta	Total
1. Number of ALCB Stores—August 1993	24	23	158	205
Number of Wine and Cold Beer Stores (pre-privatization of ALCB stores)	17	18	18	53
2. Number of Licensed Private Stores—December 1, 1995	115	100	390	605
3. Number of ALCB Stores Converted to Private	24	20	71	115
4. Number of Cities Containing ALCB Stores in August 1993	1	1	153	155
5. Number of Cities Containing Private Stores	1	1	176	178
6. Number of Cities that Contained ALCB Stores that Do Not Contain Private Stores	—	—	18	18
7. Number of Cities Containing Private Stores that Did Not Contain ALCB Stores	—	—	41	41
8. Store Sales—1992*	204,034,981	165,130,430	348,835,024	718,000,435
9. Average Sales for ALCB Stores	8,501,458	7,179,584	2,166,677	—
10. Estimated 1995 Sales per Store**	2,868,988	2,670,283	1,446,362	—
11. Store Closures Sept. 1993 to Dec. 1, 1995	6	6	16	28
*There were three stores in the rest of Alberta whose sales were included here that were not open in August 1993. **Retail sales per store are estimated as follows: take the 1992 sales proportion for the area, multiply it by the fiscal 1995 domestic and import warehouse liquor sales, add a 20 percent retail markup, and divide by the number of private retail liquor stores as of December 1, 1995.

The number of municipalities (or communities) containing private liquor stores is 178, which is an increase over the 155 municipalities or communities that contained ALCB stores. However, there are 18 municipalities or communities that contained ALCB stores that do not contain private liquor stores, and 41 municipalities or communities that contain private liquor stores that did not contain ALCB stores.

Lines 8 and 9 in table 3 contain figures on 1992 retail liquor store sales for Calgary, Edmonton, and the rest of Alberta, and the corresponding average sales for ALCB stores. In the absence of retail sales data for private liquor stores, an attempt has been made to estimate what the sales per private store might be in 1995. This was done as follows: first, we take the 1992 sales proportion for the area (like Calgary), multiply it by the fiscal 1995 (April 1, 1995 to March 31, 1996) domestic and import warehouse liquor sales, add a 20 percent retail markup, and divide by the number of private liquor stores as of December 1, 1995.¹⁷ Using this procedure, Calgary's 1995 estimated sales per store are $2,868,988, Edmonton's are $2,670,283, while the rest of Alberta has an average estimated sales per store of $1,446,362. The increased liquor store count under privatization implies much lower sales per store compared to ALCB stores, with the largest reductions occurring in Calgary and Edmonton.

Under privatization, 15 communities have experienced an increase in their liquor store count by more than a factor of four, and 11 have had their store count increase by a factor of four. The most dramatic increase occurred in Grande Prairie, where the number of liquor stores increased from one to ten. In Banff, the number of liquor stores increased from one to seven.

One interpretation of these numbers is that the ALCB, as a retail liquor monopolist, chose the profit-maximizing number of stores to serve each community. Multiple liquor stores in smaller communities would lead to higher average costs for the ALCB as scale economies at the store level would be less fully exploited. With relatively free entry under privatization, new private stores would be established in communities until potential entrants perceived that additional entry would result in losses.

Two other characteristics of the location data deserve comment: the number of store closures since privatization, and the growth of liquor store chains. Store closures have been relatively few between September 1993 and December 1, 1995. Only six stores in Calgary, six stores in Edmonton, and 16 stores in the rest of Alberta were closed as of December 1, 1995. These 28 closures represent less than five percent of the 632 private liquor stores that had been licensed as of December 1, 1995. (As of April 30, 1996, only eight stores in Calgary, 11 stores in Edmonton, and 25 stores in the rest of Alberta have been closed. These 44 closures represent 6.5 percent of the 672 private liquor stores that had been licensed as of April 30, 1996.)

With respect to retail chains, privatization under Model 1 could theoretically create incentives for the growth of liquor store chains. However, it was argued in Section 3 that the imposition of uniform wholesale prices removes much of this incentive. As of December 1, 1995, well under 10 percent of the private liquor stores in Alberta were members of chains. The low level of chain development can be interpreted as an inefficient outcome resulting from the adoption of Model 1 and its associated restrictions. A liquor store market populated by independent retailers does little to economize on consumer search costs. A liquor store chain can promote its store brand and establish a reputation for carrying a certain variety and for charging the same prices at all of its stores in a given geographic market. An increase in the proportion of chain stores in a market should reduce the amount of price dispersion, and hence the overall level of search costs.

Table 4 is designed to show how the liquor store size distribution has likely shifted under privatization. The first two columns of table 4 show the number of communities with average ALCB store sales of a given amount. (Only eight communities had more than one ALCB store.) The third column shows the number of communities with average sales per private store of a given amount. The average sales per private store in a community are calculated on the basis of 1992 ALCB store sales to the community, and the number of private stores in the community as of December 1, 1995. There are only 18 of 137 communities that used to contain ALCB stores that now contain private liquor stores that have average estimated sales per private store of more than $1.5 million. In contrast, there were 66 of 155 communities that used to contain ALCB stores that had average sales per store of more than $1.5 million.

Another way of seeing how privatization has affected the liquor store distribution is with the aid of a diagram. Figure 1 plots liquor stores on a map of Edmonton. The squiggly line running through the middle of Edmonton represents the North Saskatchewan River that divides Edmonton into north and south pieces. The 19 closed and numbered diamonds on the map are the locations of ALCB stores that are now privately operated. The three open diamonds on the map are the locations of ALCB stores that were not converted to private liquor stores. The half open, half closed diamond on the map (store 32) is the location of the one ALCB store that was converted to a private liquor store, but closed prior to December 1, 1995. The polygon within which each of the former ALCB stores is contained is the nearest point set of the store. On the assumption that all stores are identical and charge the same prices, that transportation costs are an increasing function of Euclidean distance, and that consumers are cost minimizers and minimize distance travelled, these nearest point sets represent the market areas or trade areas of the stores. (Each trade area contains all consumers living closer to it than to any other store. The boundary between two stores is the perpendicular bisector between the stores' locations.) The trade areas help to illustrate how areas once served by one ALCB store are now served by a multiplicity of private liquor stores.

Table 4: Liquor Store Size Distributions
Sales (thousands)	Number of Communities with Average ALCB Store Sales of:	Number of Communities with Average Sales per Private Store of:*
100-200	4	1
201-300	3	7
301-400	10	9
401-500	6	14
501-600	16	13
601-700	7	8
701-800	6	11
801-900	12	12
901-1,000	10	12
1,001-1,100	2	7
1,101-1,200	4	8
1,201-1,300	5	7
1,301-1,400	2	8
1,401-1,500	2	2
1,501-1,600	3	4
1,601-1,700	2	3
1,701-1,800	3	3
1,801-1,900	2	2
1,901-2,000	1	3
2,001-3,000	15	1
3,001-4,000	21	2
4,001-5,000	4
5,001-6,000	4
6,001-7,000	7
7,001-8,000	2
8,001-9,000	1
9,001-10,000	0
10,001-11,000	0
11,001+	1
*The average sales per private store in a community are calculated on the basis of 1992 ALCB store sales to the community, and the number of private stores in the community as of December 1, 1995.

The black dots on the map represent the locations of private liquor stores listed in the 1995 Edmonton Yellow Pages that are still open as of December 1, 1995. (The 1995 Edmonton Yellow Pages is based on location data that would have been effective as of October 1994, about one year after privatization.) The five open dots are the locations of private liquor stores that were listed in the 1995 Yellow Pages, but that closed prior to December 1, 1995. The Xs mark the locations of 22 private liquor stores that were not in the 1995 Yellow Pages, but that were open as of December 1, 1995.

Liquor outlet density has increased the most in the area formerly served by ALCB store 59. There are nine private stores located in that area, where formerly there was one ALCB store. The neighbouring area has an outlet density of eight, where there used to be one ALCB store. Areas with stores 24 and 46 now have three and four stores, respectively. These four ALCB store trade areas contain the "inner city," the downtown area, and many of the apartment and condominium highrises west of the downtown area. Twenty-four private liquor stores have replaced four ALCB stores in these areas.

A second part of the city where there has been a noticeable proliferation of liquor store outlets is in the areas marked by stores 44 and 42. Thirteen private liquor stores have replaced the two ALCB stores serving these areas.

Figure 2 re-plots all of the same stores as plotted on Figure 1, except for the three closed ALCB stores (marked with open diamonds on Figure 1). The market areas plotted on Figure 2 are for the numbered store locations. These are the private liquor stores that were listed in the 1995 Edmonton Yellow Pages. The figure illustrates how relatively small some of the nearest point set trade areas have become, especially in the downtown area and area just west of downtown. Even some of these plotted trade areas contain additional liquor stores (marked by Xs) that were not listed in the 1995 Edmonton Yellow Pages, but that were open as of December 1, 1995. Even so, there was only one store closure (store 28) in the downtown area between September 1993 and December 1, 1995. Figure 2 also illustrates how dispersed the private liquor store locations have become. The lower density of liquor stores on the periphery of the city reflects the lower population density on the urban periphery.

With respect to chain store locations, two of the chains have adopted dispersed locations and their stores do not share market area boundaries. One of these chains, Liquor Barn, has stores that are numbered 30-34, while the other, Liquor Stop, has stores numbered 41-45. The third chain, Liquor Depot, has stores numbered 35-38. Stores 36 and 37 are neighbours, and so are stores 36 and 38 (although the latter stores are separated by the North Saskatchewan River). With the small number of dispersed locations occupied by chain stores, one would have difficulty finding any locational evidence to support a finding of strategic locational behaviour on the part of liquor store chains.

Finally, Calgary store locations need to be discussed. Calgary's store locations could have been plotted in the same way as Edmonton's and similar results would have been obtained. In Calgary, there were 24 ALCB stores in August 1993. In the October 1995 Calgary Yellow Pages (which probably contains liquor store locations as of June 1995), there were 102 liquor stores listed. Four of these had closed by December 1, 1995, while an additional 17 stores had opened, for a total of 115 liquor stores by December 1, 1995. Liquor store locations are dispersed, and liquor store chains also have dispersed stores. Only the Royal Liquor Merchants chain has stores that are relatively close to one another, but there are only three of them.

To sum up our findings with respect to liquor store locations, there has been a 134 percent increase in the number of liquor stores in Alberta (as of December 1, 1995) over the combined number of beer, wine, and ALCB stores in August 1993. Most of the ALCB stores in Calgary and Edmonton were converted to private liquor stores, but less than half of the ALCB stores in the rest of Alberta were converted to private liquor stores. The number of communities served by liquor stores has increased under privatization. There have been relatively few closures of private liquor stores between September 1993 and December 1995, and this might be partly explained by the fact that the growth of retail liquor store chains has been slow. Less than 10 percent of private liquor stores in Alberta are members of chains. Liquor store chains have generally chosen to locate in a dispersed fashion within communities, so it would be difficult to find any locational evidence to support a finding of strategic locational behaviour on the part of liquor store chains.

Product selection

With respect to product selection, one would expect a government owned liquor store monopolist to choose product selection to maximize profits. Under privatization Model 1, product selection becomes a vehicle for non-price competition between stores. Increasing product selection at a given store should increase the demand enjoyed by the store since more selection reduces a consumer's search costs. (It does so by increasing the probability that consumers will find their most desired products when they visit the store.) In addition, stores will have an incentive under privatization to carry a larger selection than under government ownership because a store that does not stock a consumer's most desired product may not be chosen on a subsequent liquor shopping trip—something that a government owned store would not be concerned about. Hence, one might expect product selection to increase on average as a result of privatization.

Section 2 reported that the product selection at ALCB stores varied depending on whether the store was an A store (with 600-700 stock keeping units (or SKUs)), a B store (with 1100 SKUs), a regular C store (with 1500-1600 SKUs), or an expanded specialty C store (with 2600 SKUs). A given brand that comes in, say, five different package sizes would have five different CSPC numbers and be counted as five SKUs. With information on the store type for each ALCB store, one can calculate the weighted average number of SKUs of ALCB stores by area; the results appear at the bottom of table 5. Calgary ALCB stores had an average product selection of 1,369 SKUs, Edmonton ALCB stores had an average of 1,380 SKUs, and the rest of the province's ALCB stores averaged 824 SKUs. The provincial average product selection for ALCB stores is 950.

In order to compare the ALCB store product selection with the private liquor store product selection, survey data on product selection in Alberta liquor stores are required. Westridge Marketing Services carried out a survey of product selection in 100 liquor stores in Alberta in February 1996. Twenty-eight of the surveyed stores were located in Edmonton, 28 were located in Calgary and 44 were located in a variety of smaller communities across Alberta. Product selection data were collected by product type, and the results of the survey are reported in table 5. For the province as a whole, the average number of SKUs per private liquor store was 1,052, which is higher than the weighted average number of SKUs per ALCB store (i.e., 950). Table 5's next three columns look at the average numbers of SKUs in private liquor stores in Calgary, Edmonton, and the rest of Alberta. For Calgary, the average number of SKUs was 1,284, while for Edmonton, the average number of SKUs was 1,142. While these numbers are less than the average product selection available at ALCB stores in Calgary and Edmonton, there were a number of sample stores in Calgary (four) and Edmonton (six) with a product selection that exceeded the 1,600 SKUs of an ALCB "C" store. For the rest of Alberta, the average number of SKUs was 847, and this figure exceeds the weighted average product selection available at ALCB stores in the rest of Alberta. The product selection range in the 100 store sample is from 183 SKUs in a small rural store to 4,191 SKUs in a Calgary store. The latter figure substantially exceeds the 2,600 SKUs available in an expanded specialty ALCB store.

Table 5: Product Selection in Retail Liquor Stores and ALCB Warehouse Pre- and Post-Privatizationa
	Provincial Average (n=100)	Calgary Average (n=28)	Edmonton Average (n=28)	Rest of Province Average (n=44)	St. Albert Warehouse SKUs (Oct 1993)b	St. Albert Warehouse SKUs (Dec. 1994)c	St. Albert Wareouse SKUs (Dec 1995)c
Beer	132	144	139	119	98	310	429
Wine	431	588	468	307	1,238	2,521	2,713
Canadian Whisky	80	83	85	74	90	117	147
Scotch	38	46	45	28	58	77	115
Vodka	56	63	64	47	79	114	145
Rum	74	83	87	59	84	133	169
Brandy/Cognac	24	30	26	19	47	84	111
Gin, Tequila, Other Spirits	30	36	29	27	62	105	173
Liqueur	90	94	99	82	134	225	319
Coolers, Cider	82	92	90	71	55	110	149
Other	16	25	8	15	12	61	43
All Products	1,052	1,284	1,142	847	1,957	3,857	4,513
Pre-Privatization ALCB Store Product Selection (Weighted Average)	Provincial Average (n=205) 950	Calgary Average (n=24) 1,369	Edmonton Average (n=23) 1,380	Rest of Province Average (n=158) 824
aPost-privatization figures are derived from the February 1996 liquor store survey carried out by Westridge Marketing Services. The product selection range in the 100 store sample is from 183 in a small rural store to 4,191 in a large urban store. bThese figures are derived from the ALCB stock catalogue and do not include SKUs carried in the warehouse as part of the Agent’s Listing Program. cThese figures are derived from the ALCB’s Liquor Wholesale Price List and do not include limited supply products available in the ALCB warehouse but not listed on the wholesale price list.

Looking at product selection data by category in table 5, it is evident that, on average, product selection in Calgary and Edmonton stores is greater than that in stores in the rest of Alberta in every product category. It is also evident that the difference in product selection between stores in Calgary-Edmonton and in the rest of Alberta is largely due to the greater variety of wine available in Calgary-Edmonton liquor stores.

The ALCB has provided some product selection figures for a few stores in Edmonton and Calgary and for five liquor store chains.¹⁸ One chain was reported to have 3 ,900 SKUs per store, while the other four chains reportedly ranged from 1,400 SKUs to 2,000 SKUs per store. The chains have most of their stores in Calgary and Edmonton. Product selection for one independent Edmonton store and five independent Calgary stores reportedly ranged from 2,500 to 3,500 SKUs. It seems clear that at least some private liquor stores provide a product selection that surpasses the 2,600 SKUs of the ALCB's expanded specialty stores (of which there was one in Calgary and one in Edmonton).

The final measure of product selection that can be examined relates to the number of SKUs available at the St. Albert ALCB warehouse. At the time of privatization, there were 1,957 SKUs listed in the general stock, specialty stock, and expanded specialty stock catalogues and carried in the ALCB warehouse (see table 5).¹⁹ There were also reported to be 1,221 SKUs carried in the warehouse as part of the Agent's Listing Program that were not listed in the stock catalogues. By December 1994 the number of SKUs in the ALCB warehouse had increased to 3,857. By December 1995 the number of SKUs in the ALCB warehouse had increased further to 4,513.²⁰ Some product categories, like wine, scotch, and rum, had double the SKUs in December 1995 compared to October 1993. Other product categories, like beer, gin/tequila/other spirits, liqueur, and coolers/cider, had substantially more than double (quadruple in the case of beer) the SKUs in December 1995 compared to October 1993.

As mentioned previously, all of the products in the warehouse are now on consignment from suppliers. Suppliers make the decisions regarding which products to bring into the warehouse, and bear the handling and storage costs of poor decisions. Suppliers, freed from the entry and listing restrictions imposed on them by the ALCB, have greatly increased the variety of products available to retailers. Retailers have found it in their interest to engage in non-price competition by increasing the selection of products available to consumers.

Is the increase in selection due to the proliferation of brands in particular product classes, or an increase in the variety of package sizes? Consider Canadian whisky first. In October 1993, there were 35 different Canadian whisky brands and 90 SKUs represented by 10 agents. In December 1995, there were 43 different Canadian whisky brands (a 23 percent increase) and 147 SKUs (a 63 percent increase), represented by 15 agents (a 50 percent increase).

Next, consider Canadian beer. In October 1993, there were 42 brands of Canadian beer and 70 SKUs, produced by six different breweries. In December 1995, there were 114 different brands of Canadian beer (a 171 percent increase), and 257 SKUs (a 267 percent increase), produced by 20 different breweries (a 233 percent increase).

It is clear that in the cases of Canadian whisky and Canadian beer, the number of brands, number of package sizes, and number of producers listing products on the wholesale price list have all increased since privatization. In percentage terms, the variety of package sizes carried in the warehouse has increased more than the number of brands.

In terms of the impact that privatization has had on product selection, then, for a sample from 100 private liquor stores across Alberta, the average number of SKUs per private liquor store is higher than the weighted average number of SKUs per ALCB store. Product selection in Calgary and Edmonton stores is greater than in stores in the rest of Alberta in every product category, but particularly in the wine category. The number of SKUs available at the ALCB warehouse has more than doubled under privatization, and this increase reflects both an increase in the number of brands in particular product classes, and an increase in the variety of package sizes.

Price

It can be argued that liquor store privatization will not necessarily result in significant changes in retail prices or government revenues if the government's objective with respect to the sale of liquor products both before and after privatization is net revenue maximization. However, some changes in retail price would be expected to the extent that privatization resulted in 1) lower retailer costs (from the expansion of retail chains or lower labour costs for example), 2) higher supplier costs (from the higher warehouse storage and handling costs and delivery costs from the warehouse, the higher marketing costs incurred by trying to sell products to the large number of private liquor stores, or a non-revenue neutral change in the government's markup), 3) more intense spatial competition due to an increasing number of stores serving the market, and 4) moving from a market in which all stores charge the same price to one in which there will be price dispersion due to imperfect information and costly consumer searching. We have already seen that the expansion of retail liquor store chains in Alberta has been quite modest, and so overall efficiencies resulting from chain formation are not expected to be large. The other sources of price change after privatization cannot so easily be discounted.

To assess the impact of privatization on retail prices of liquor products, we use survey price data contained in a Retail Price Survey for Alberta Liquor Stores. The survey is carried out each month by Westridge Marketing Services. In the survey dated January 15, 1996, data on 187 product prices were collected from 100 private liquor stores in Alberta. Twenty-eight stores are sampled in both Calgary and Edmonton (seven in each of four quadrants of each city), and 44 in the rest of the province. The average price in each city quadrant or community is reported for each product, along with the highest and lowest sampled price, provincial average retail price, wholesale price, and average markup on wholesale. Many of the products included in each month's price survey have been chosen at the request of specific suppliers. Still, each product category is represented in the survey, and many well known, high volume products are also included. The Retail Price Survey carried out by Westridge is the most comprehensive survey of liquor prices available for Alberta, and the fact that a similar price survey was conducted in January 1995 means that price comparisons over time are possible.

Table 6 reports the results from calculating the price changes of products listed in the January 1996 Retail Price Survey and the October 1993 ALCB General Stock Catalogue. The private liquor store prices used are the provincial average prices reported in the Retail Price Survey. Some product categories, like Canadian whisky, vodka, and beer, are reasonably represented in the price survey, while others, like wine and coolers, are not. (See the last column of table 6 for the number of SKUs in the ALCB warehouse in December 1995, by product category.) Looking at the average percentage change in price column, the overall average increase in price for 143 products from October 1993 to January 1996 is calculated to be 8.46 percent. There is quite a lot of variation in the average percentage change in price across product categories. Red and white wine have the largest percentage price increases (at 15.39 and 14.53 percent, respectively) but there were only 17 of these wine products included in the price survey. Closer to the average percentage price change are Canadian whisky (7.84 percent), vodka (8.36 percent), and beer (9.21 percent).

Table 6: Liquor Product Price Changes Between October 1993 and January 1996*
Product Category	Number of Products	Average Percentage Change in Price	Standard Deviation	Number of Positive Price Changes	Number of Negative Price Changes	Number of Unchanged Prices	Minimum Percentage Change in Price	Maximum Percentage Change in Price	Total SKUs in ALCB St. Albert Warehouse
Canadian Whisky	29	7.84	2.73	29	0	0	1.44	12.79	147
Scotch	8	.19	5.55	4	4	0	-6.17	10.08	115
Vodka	16	8.36	3.72	16	0	0	2.13	14.76	145
Rum	12	11.80	1.52	12	0	0	9.84	14.46	169
Other Spirits/ Liqueurs	10	6.25	12.72	6	3	1	-21.90	18.27	559
White Wine	10	14.53	6.73	10	0	0	4.58	24.41	1,265
Red Wine	7	15.39	6.50	7	0	0	1.35	24.38	1,092
Other Wine	5	-2.64	16.26	3	2	0	-29.59	13.42	202
Coolers	3	2.60	.99	3	0	0	1.25	3.61	113
Beer	43	9.21	4.37	43	0	0	.06	25.81	429
All	143	8.46	7.20	133	9	1	-29.59	25.81	4,236
All (Excluding wine and some beer products)	115	8.01	5.89	107	9	1	-29.59	25.81	—
*The private liquor store prices used are the provincial average prices reported in the Retail Price Survey carried out by Westridge Marketing Services.

Three of the columns in table 6 count up the direction of price change for the various products. In total, 133 products had price increases from October 1993 to January 1996, nine had price reductions, and one had an unchanged price. The largest percentage price reduction was 29.59 percent for a champagne product while the largest percentage price increase was 25.81 percent for a beer product.

By way of comparison, the Consumer Price Index for Alberta increased about 4.1 percent from October 1993 to January 1996, while the Consumer Price Index for Canada increased about 2.4 percent. In real terms, the average percentage price increase was about half the nominal increase of 8.46 percent.²¹ When discussing price changes, one should also take into account the average consumer's lower transportation costs brought about by the increase in the number of liquor stores under privatization. Everything else being equal, lower transportation costs under privatization would have lowered the delivered price of the product, which is the sum of the price at the store and transportation costs. So any increase in the price of a liquor product at the store will be offset to some extent by a reduction in transportation costs for many consumers.

Another approach to summarizing the price changes is to calculate price indices for the liquor products contained in the price survey. We can calculate both the Laspeyres price index, which uses base period quantities as weights, and can be written as L_p = , and the Paasche price index, which can be written as P_p = and which uses the given period's quantities as weights. ²² With October 1993 as the base period, and January 1995 as the given period, the Laspeyres price index is 1.0857. With 1993 as the base period, and January 1996 as the given period, the Laspeyres price index is 1.0990. Thus, the Laspeyres price index indicates almost a 10 percent price increase from October 1993 to January 1996, which is slightly higher than the 8.46 percent price increase obtained by simply averaging price changes.

With October 1993 as the base period, and January 1996 as the given period, the Paasche price index is 1.0904, and with January 1995 as the base period and January 1996 as the given period, the Paasche price index is 1.0093. The Paasche price index indicates about a 9 percent increase in price for October 1993 to January 1996. Both the Laspeyres and Paasche price indices indicate that most of the price increase between October 1993 and January 1996 occurred by January 1995.

One possible source of the retail price change from October 1993 to January 1996 could be an increase in wholesale prices, brought about by supplier price increases, increases in storage and handling costs, or the implementation of the flat markup. Table 7 provides calculations of wholesale price changes from November 1993 to December 1995 for products that were contained in both the January 1995 and January 1996 Retail Price Surveys. (Wine products have been excluded.) In all product categories, the average percentage change in wholesale price is negative. Beer had the smallest percentage price change (i.e., -.11 percent), while vodka had the largest (-9.27 percent). Over all 115 products for which the calculation was made, the average percentage wholesale price reduction was 3.37 percent. The last line of table 6 shows that the average increase in the retail price for the same set of products was 8.01 percent from October 1993 to January 1996. One possible explanation for these results is that retailers have increased their profit margins and markups over wholesale prices between October 1993 and January 1996. (Some small increase in markup of retail price over wholesale price is evident from data contained in the January 1995 and January 1996 Retail Price Surveys.) The other possible explanation for the result is that retail prices could have increased by more than 8 percent on average at the time of privatization, so that wholesale price reductions from November 1993 to December 1995 could have been met with retail price reductions and still leave average prices higher than they were just before privatization.

Table 7: Changes in Wholesale Prices: November 1993 to December 1995
Category	Number of Products	Average Percentage Change in Price	Standard Deviation	Number of Positive Price Changes	Number of Negative Price Changes	Number of Unchanged Prices	Minimum Percentage Change in Price	Maximum Percentage Change in Price
Canadian Whisky	29	-4.40	2.50	0	29	0	-11.53	-1.18
Scotch	8	-2.46	5.68	3	5	0	-7.86	6.06
Vodka	16	-9.27	4.23	0	16	0	-13.88	-.83
Rum	12	-3.11	2.17	0	12	0	-9.01	-.56
Other Spirits/ Liqueurs	10	-1.70	2.87	3	7	0	-7.33	2.35
Coolers	3	-10.10	1.15	0	3	0	-11.71	-9.15
Beer	37	-.11	4.53	17	19	1	-10.03	11.68
All	115	-3.37	4.90	23	91	1	-13.88	11.68

Other results reported in table 7 are that there were almost four times as many wholesale price reductions as price increases over the November 1993-December 1995 period, with beer products having almost as many wholesale price increases as price reductions. Over all products, the largest wholesale price reduction is 13.88 percent, while the largest wholesale price increase is 11.68 percent.

Given imperfect information regarding retail liquor store prices, and the fact that consumers have to engage in costly searches to learn about prices, one would expect a significant degree of retail price dispersion within cities and towns (see Carlson and McAfee (1983) and Dahlby and West (1986)). One measure of dispersion is the coefficient of variation, which can be calculated by dividing the standard deviation by the mean. The January 1996 Retail Price Survey reports the average price for sample stores in each of eight quadrants that make up Calgary and Edmonton (four quadrants in each city), and the average price for sample stores in 24 other communities in the rest of Alberta. The coefficient of variation can be calculated using these 32 observations on average prices, and for a subset of 67 products, the average coefficient of variation is calculated to be 2.77 percent. The individual coefficients of variation ranged from 1.34 percent to 8.76 percent. Given that these coefficients of variation have been calculated from observations of average prices in a sample of communities, the actual extent of price dispersion is higher than that reflected in the reported coefficient of variation statistics. Substantial price dispersion would appear to characterize retail liquor prices in Alberta.

Table 8: Schedule of Payments to Governments (in thousands)
	Fiscal Year End March 31, 1996 (52 weeks)	Fiscal Year End March 31, 1995 (64 weeks)	Fiscal Year End January 4, 1994 (52 weeks)	Fiscal Year End January 5, 1993 (52 weeks)	Fiscal Year End January  7, 1992 (52 weeks)
Federal Government
Excise and Duty Taxes	$79,031	$106,429	$79,816	$83,908	$85,373
Goods and Services Tax	66,790	92,079	71,772	67,710	60,205
Total	145,821	198,508	151,588	151,618	145,578
Remittances to Provincial Government	485,000	483,000	454,500	434,500	439,000
(Profit on Sales)a	410,458	525,157	423,599	402,779	405,727
Municipal Governments
Property Taxes	1,001	1,965	3,179	3,123	3,137
Business Taxes	28	90	617	543	786
Total	1,029	2,055	3,796	3,666	3,923
Payments to Governments by ALCB	631,850	683,563	609,884	589,784	588,501
Alberta Brewers excise taxes paid to federal government	46,768	65,694	51,723	51,377	45,399
Total Payments to Governments	$678,618	$749,257	$661,607	$641,161	$633,900
aProfit on sales is calculated as Sales – Cost of Goods Sold – Operating Expenses. Source: Sixty-ninth, Seventieth, Seventy-first, and Seventy-second Annual Reports of the ALCB.

Results on changes in retail liquor prices since privatization suggest that in spite of any lower private retailer operating costs that could have been achieved by reducing labour costs, for example, and in spite of more intense spatial competition among liquor retailers and evidence of falling wholesale liquor prices (on average), average nominal retail liquor prices have increased since privatization. Assuming that the government was already charging monopoly prices for liquor products prior to privatization, one would not expect the move under privatization to a market where there is price dispersion due to imperfect information, and high consumer search costs to lead to an increase in price. However, it is possible that part of the explanation for increasing retail liquor prices could be a non-revenue neutral change in the provincial government's markup accompanying privatization. The next section will examine the Alberta government's liquor revenue figures.

Government revenues

If the government's policy with respect to the sale of liquor products is net revenue maximization both before and after privatization, then it will wish to choose its liquor price markup after privatization to achieve that objective. The implication here is that the government should seek to implement a revenue neutral markup and that goal was in fact the Alberta government's stated objective at the time that the flat markup was introduced. (Recall that in economic terms, the flat markup is like a specific tax imposed at the wholesale level so that it is one component of the wholesale price.)

Table 8 contains the schedule of payments to governments for the last five fiscal years. The payments to the federal government are relatively flat for the 1991-1993 fiscal years, and would have increased for the 1994 fiscal year if we adjust the reported figure to reflect a 52-week (instead of 64-week) fiscal year. Remittances by the ALCB to the Alberta government increased from 1991 to 1993. The figure for the 1994 fiscal year shows remittances to the provincial government increasing, but when adjusted to a 52-week basis, the figure would show remittances falling. This is partly due to the fact that the ALCB began the fiscal year with $100 million in remittances in excess of net unappropriated income.

What is perhaps more revealing is the ALCB's profit on sales, reported in table 8. (Profit is calculated as Sales – Cost of Goods Sold – Operating Expenses.) Profits fell marginally between 1991 and 1992, but increased by 5.2 percent in 1993—the year in which privatization occurred. Profits increased marginally again in 1994 when the profit figure is adjusted to a 52-week basis. The Alberta government reduced its flat markup rates in August 1994 and January 1996 in an effort to maintain revenue neutrality. The ALCB's profit on sales reported for the fiscal year ending March 31, 1996 has, in fact, fallen to a level just two percent above what it earned in the last pre-privatization fiscal year of 1992. There is thus some evidence to support the hypothesis that the change from an ad valorem to a flat markup at the time of privatization was not revenue neutral, but rather led to a modest increase in government revenues. This increase could at least partly explain the observed increase in the average retail liquor price under privatization.

In choosing its markup, the Alberta government must take into account the retail liquor prices in neighbouring jurisdictions. If Alberta's retail liquor prices are significantly higher than those in neighbouring jurisdictions, one would expect some Alberta consumers (particularly those living near provincial borders) to shop for liquor in other jurisdictions. One might also observe an increase in illegal imports and sales of liquor products in Alberta. We have no estimate of the magnitude of cross-border shopping for liquor, nor the extent of smuggling activity as it relates to liquor products. We do, however, have data on the quantities of liquor products sold (by category) in Alberta; these figures appear in table 9. Figures for fiscal year 1994 were converted to a 52-week basis by multiplying them by 52/64.

Table 9: Quantities (Hectolitres) of Liquor Products Sold by Category
	Fiscal Year
	1995 (52 weeks)	Percentage Change 1994-95	1994 (64 weeks)	1994* (52-week basis)	Percentage Change 1993-94	1993 (52 weeks)	Percentage Change** 1992-93	1992 (52 weeks)	Percentage Change 1991-92	1991 (52 weeks)
Spirits
Whisky	62,916	-9.12	85,208	69,231	-4.52	72,508	4.22	69,569	-7.16	74,937
Gin	3,387	-6.72	4,469	3,631	-7.06	3,907	3.22	3,785	-11.15	4,260
Liqueurs	18,899	2.22	22,756	18,489	-1.98	18,863	5.17	17,935	-4.92	18,863
Rum	30,828	-0.89	38,282	31,104	-5.59	32,946	7.38	30,682	-2.96	31,617
Vodka	38,272	-0.81	47,490	38,586	-2.13	39,424	4.90	37,582	-5.99	39,976
Others	8,946	113.10	5,167	4,198	-3.03	4,329	-0.41	4,347	3.87	4,185
Wine
Dessert	8,031	-13.14	11,380	9,246	-7.56	10,002	—	7,868	-5.52	8,328
Table/ Sparkling	144,885	-1.41	180,871	146,958	-3.49	152,269	—	144,959	-5.29	153,062
Other	—	—	—	—	—	—	—	3,150	-2.26	3,223
Coolers/ Ciders	45,805	32.11	42,673	34,672	0.22	34,597	-1.33	35,062	-6.20	37,379
Beer
Packaged	1,524,675	2.88	1,824,077	1,482,063	-0.32	1,486,837	2.74	1,447,249	-4.33	1,512,777
Draft	312,618	0.57	386,954	314,400	-6.85	337,527	3.36	326,568	6.25	307,358

From 1991 to 1992, the quantities of liquor products sold fell in every category except other spirits and draft beer. From 1992 to 1993, the quantities of liquor products sold increased in every category except other spirits and coolers/ciders. Part of this increase might be attributed to the opening of private liquor stores in the last quarter of 1993, and the need to stock their shelves. From 1993 to 1994 (adjusted), the quantities of liquor product sold fell in every category except coolers/ciders. However, the drop in quantities sold, which is not large, is likely overstated by the crude correction used to put fiscal 1994 on a 52-week basis. The quantities of product sold during the last quarter of 1994, which includes the Christmas season, likely exceed those sold in the first quarter of 1995. Comparing quantities sold in 1995, which is reported on a 52-week basis, with quantities sold in 1993 and 1992, one finds that quantities of spirits sold dropped by 5 percent and less than 1 percent respectively. Quantities of wine sold were off 6 percent in 1995 compared to 1993, and down 2 percent in 1995 compared to 1992. Cooler sales were up close to a third in 1995 compared to 1993 and 1992, reflecting their increased popularity. Packaged beer sales were up 2.5 percent in 1995 compared to 1993, and up 5.3 percent in 1995 compared to 1992.

Table 10: Retail Liquor Price Comparisons for B.C., Saskatchewan, and Ontario with Alberta—January 1996
Product Category	Alberta - B.C.		Alberta - Saskatchewan		Alberta - Ontario
	Number of Products in BC-Alberta Comparison	PBC-PAB PAB	Number of Products in Saskatchewan - Alberta Comparison	PSK-PAB PAB	Number of Products in Ontario- Alberta Comparison	PON-PAB PAB
Canadian Whisky	26	-10.19	26	-3.78	25	-7.77
Scotch	8	7.05	8	8.98	8	0.16
Vodka	15	-8.92	15	-3.52	12	-3.91
Rum	12	-8.77	12	-4.88	10	-6.50
Other Spirits/Liqueurs	9	3.87	10	1.70	10	1.59
White Wine	10	3.92	8	-1.43	4	-5.55
Red Wine	11	6.22	7	-1.03	4	-6.51
Other Wine	4	0.78	5	7.43	5	-2.39
Coolers	2	-1.00	5	27.26	4	3.53
Beer	42	-1.45	2	10.98	40	2.90
All	139	-2.63	98	0.57	122	-1.80

The figures in table 9 suggest that there has not been either a large shift upward or downward in liquor products sold since privatization was announced in September 1993. Hence, cross-border shopping and smuggling are unlikely to be significantly greater problems after privatization than they were before privatization.²³ The figures in table 9 also suggest that retail price differences between Alberta and neighbouring jurisdictions are not so large as to induce cross-border shopping and smuggling behaviour. This prediction is confirmed in a comparison of a sample of retail liquor prices in B.C., Saskatchewan, and Ontario with the average prices of the same products in Alberta. The results of the comparison are reported in table 10. The Alberta prices used in the comparison are the provincial average prices taken from the January 1996 100-store retail price survey carried out by Westridge Marketing Services. (Recall that 28 of the surveyed stores are in Calgary, 28 are in Edmonton, and the other 44 are in smaller communities across Alberta.) Prices for B.C. are taken from the B.C. Liquor Distribution Branch General Products Price List for the period December 31, 1995 to January 27, 1996. Saskatchewan's liquor prices are reported in the February 5, 1996 issue of the Saskatchewan Liquor and Gaming Authority Official Price List. Ontario's liquor prices are contained in the Liquor Control Board of Ontario's Master Brand List, January 29, 1996. All prices used in the comparison include provincial taxes and GST.

Looking first at the Alberta-B.C. comparison, one finds that in five of the product categories, on average, prices in B.C. exceed the average prices in Alberta, while in five of the product categories the reverse is true. The overall average percentage by which Alberta prices exceed B.C. prices is 2.63 percent. It is important to keep in mind, however, that while some average Alberta prices are higher than those in B.C., there is substantial retail price dispersion in Alberta. It will frequently be possible for a consumer to find a lower product price in Alberta than in B.C. provided the consumer shops around.

Looking next at the Alberta-Saskatchewan comparison, one finds that in five product categories, on average, prices in Saskatchewan exceed the average prices in Alberta, while in five of the product categories the reverse is true. Note that only two beer products are contained in the Alberta-Saskatchewan price comparison because the Saskatchewan price list does not contain the price of Canadian beer products. With that limitation in mind, the overall average percentage by which Saskatchewan prices exceed Alberta prices is 0.57 percent.

Finally, consider the Alberta-Ontario comparison. On average, prices in Ontario exceed the average prices in Alberta in four product categories, while the reverse is true in six product categories. The overall average percentage by which Alberta prices exceed Ontario prices is 1.80 percent.

Given the small size of the average percentage differences between Alberta's and B.C.'s and Saskatchewan's liquor prices, and given that most consumers do not live close to the B.C. or Saskatchewan borders, most consumers will not find it convenient or rewarding to compare B.C.'s and Saskatchewan's retail liquor prices with those of stores in their community in order to save a dollar on a bottle of whisky or a case of beer. Hence, cross-border shopping for liquor products is unlikely to be quantitatively significant or to have a significant adverse impact on government revenues given current price differentials between Alberta and its neighbours. The only exceptions to this conclusion would be those few liquor products whose price differentials are high enough to make incurring the transactions costs of moving liquor products across provincial borders worthwhile. In the price comparison survey data, the largest provincial percentage price differences were observed for a liqueur product and a cognac product (with prices lower in Alberta). These price differences likely result from Alberta's change to a flat markup from an ad valorem markup after privatization.

To summarize, Alberta government liquor revenues have not been adversely affected by privatization and the shift from government liquor revenues based on a liquor store markup to revenues derived from a flat markup imposed at the wholesale level. The data also indicate that there has not been either a large shift upward or downward in the quantity of liquor products sold since privatization was announced in September 1993. In addition, a comparison of a sample of Alberta's average liquor product prices with the prices of the same products in each of B.C., Saskatchewan, and Ontario, reveals that the average percentage price differences are relatively small.

Table 11a: Employment in ALCB Stores
ALCB Store Type	1992 Sales	Mana- gers	Full-time Employ- ees	Part-time Employ- ees	Casual Employ- ees	Number of Stores in Edmonton	Number of Stores in Calgary	Number of Stores in Rest of Province
.  A	$500,000	1			2	A stores = 0	A stores = 0	A stores = 107
.  A	800,000	1		1	2
.  A	1,500,000	1		1	1
.  B	3,500,000	1	2	1	1	B stores = 11	B stores = 12	B stores = 41
.  B	7,500,000	1	4	3	6
.  C	11,000,000	2	5	6	2	C stores = 12	C stores = 12	C stores = 10

Table 11b: Employment in Private Liquor Stores
Feb 1996 Liquor Store Survey	Average Full-time Employees Per Store	Average Part-time Employees Per Store	Number of Licensed Stores (12/95)
Province (100 store sample)	2.92	4.59	605
Calgary (28 stores)	3.54	5.43	115
Edmonton (28 stores)	3.25	5.46	100
Rest of Province (44 stores)	2.32	3.50	390
Range in Feb 1996 Liquor Store Survey: 1 full-time employee and 2 part-time employees in a small rural store; 16 full-time employees and 20 part-time employees in a large urban store.

Employment and wages

The final two economic impacts of privatization that need to be examined are the impacts on employment and wages. Table 11 provides a comparison of employment in ALCB and private liquor stores. The top half of the table was constructed in the following way: six stores were selected at different points in the ALCB store size distribution and the ALCB was asked to supply employment information for these stores for a date just prior to privatization.²⁴ The table shows that all six ALCB stores had managers, but the smaller type A stores (those with a smaller number of SKUs and lower sales) operated with part-time and casual employees. All A stores are located outside Calgary and Edmonton. Type B and C stores employ full-time as well as part-time and casual employees and are found in both Calgary and Edmonton and the rest of Alberta. At the time of privatization, there were 1,394 people employed in ALCB stores in Alberta (and this represented about 950 full-time equivalents (FTEs) according to Westridge Marketing Services).

In order to get a snapshot of current employment and wages in the retail liquor store industry in Alberta, Westridge Marketing Services included questions regarding part-time and full-time employment and the average non-management wage on the same survey of 100 liquor stores in which product selection data were collected. The results of the survey appear in the lower half of table 11. For the province as a whole, the average number of full-time employees per store is 2.92, which is a number similar to that of a smaller type B ALCB store. The average number of part-time employees is 4.59, which would probably be similar to a medium-sized type B ALCB store. Looking at the employment figures for Calgary and Edmonton, they are larger than the provincial average, reflecting the larger average store size in those cities.

Extrapolating from the survey data, there were an estimated 1,637 full-time employees and 2,535 part-time employees in private liquor stores as of February 1996, for a total employee count of 4,172. (Westridge Marketing Services estimated that there were 1,600 full-time and 2,600 part-time employees in private liquor stores as of February 1996, for a total employee count of 4,200.) Assuming two part-time employees equal one full-time equivalent employee, there were an estimated 2,904 FTEs (or 3,000 FTEs based on Westridge Marketing's estimate) employed in private liquor stores. Liquor store employment has approximately tripled since privatization, as has the number of liquor stores.

Table 12a: Wages in ALCB Stores
	Range
Liquor Store Managers	$30,296 to $49,173
Warehouse Workers	$23,777 to $30,169
Liquor Store Clerks (Full-time)	$23,865 to $28,174 or $12.19/hour to $14.39/hour
Liquor Store Clerks (Casual)	$8.50/hour

Table 12b: Wages in Private Liquor Stores
Feb 1996 Liquor Store Survey	Average Wage: Non-management employees
Province (100 stores)	.    $7.19
Calgary (28 stores)	.    $7.32
Edmonton (28 stores)	.    $7.15
Rest of Province (44 stores)	.    $7.13

Along with the increase in liquor store employment has come a reduction in liquor store non-management employee wages. Table 12 shows the salary range for liquor store managers, warehouse workers, and liquor store clerks (both full-time and casual) under the ALCB and the average wage for non-management liquor store employees based on results for the February 1996 liquor store survey. For the province as a whole, the average wage paid by private liquor stores is half that paid by the ALCB to a full-time liquor store clerk at the top of the scale. The range of private liquor store wages reported was from $5.00 to $10.00 per hour.

One might conclude from these figures that the expanded network of liquor stores in Alberta has been facilitated by sharp reductions in liquor store wages. Under the ALCB, the government was the residual claimant of liquor store net revenues. The government apparently was prepared to share its liquor store revenues by paying union workers higher wages than the private sector is prepared to pay its non-union workers.

To summarize, full-time equivalent employment in liquor stores is estimated to have approximately tripled under privatization, but the wages of non-management liquor store employees are almost one-half of what a full-time union worker at the top of the scale could earn in an ALCB store.

[Previous] [Contents] [Next]

info@fraserinstitute.ca You can contact us at the above email address for any comments or information requests. Please report any dead links or technical problems.

If you know someone who would be interested in this web page, please enter their email address below, and we will forward this URL to them: Email Address:

Last Modified: January 21, 2001.