In brief, this first approach used by Professors Haring and Yoder gives no evidence of higher prices in stamp stores; in fact, if it suggests anything, it indicates just the reverse.

This is done in this chart on my right. Notice that, if you will, split the entire group of 28 stores in 2, forgetting these 2 over here for a moment, the entire 28 stores, and you will notice that of these stores and the box figures represent those which gave trading stamps-that they range the whole gamut, that 7 of the 9 stampgiving stores fall in the lowest half of all of the stores investigated in this city; 2 in the upper. If you take the average of all of the prices of the stamp-giving stores and of the non-stamp-giving stores, actually the stamp-giving stores averages a bit below I percent that of the nonstamp stores, but the significant thing is that this is the freedom of choice that competitive markets give to their customersa customer has a freedom of choice of shops at any store and prices vary from store to store. In some places the stamp-giving stores are below; other places they are not, although on the average they were below.

That approach is not new to this committee now.

Now we turn to the second approach used in the Indiana study of the question: What is the impact of trading stamps on retail prices? As basic data for this approach, the authors used retail prices as collected monthly by the Bureau of Labor Statistics for 20 cities.

The 20 cities were divided into 2 groups, a group of 15 cities in which trading stamps were widely used and a 5-city group where trading stamps were not important in the food trade. Then average monthly price trends were compared for these two groups of cities for the January 1950 to December 1956 period. That takes us back before we had food stamps being used in our major supermarkets. You will recall that it was during the last 3 or 4 years of this period that stamps were expanding rapidly in the food field.

It should be self-evident that if trading stamps allowed the retailers using them to raise prices, we would expect the price trend in the 15 stamp cities to rise relative to that in the 5 nonstamp cities. But exhibit ÎI makes it clear that this did not happen.

The 100-percent line across this chart represents the 5 nonstamp cities. That represents 100 percent all the way. The dotted line represents the average for 15 stamp cities. This goes back to the period before there were stamps. It goes through to the end of 1956 when stamps were widespread in the 15 stamp cities.

Notice that the average prices in the stamp cities were at approximately the same level at the end of this period as at the beginning. At times these would be below. You always get fluctuations. In fact, they ended up slightly below.

I think that the crucial point about this is-remember that stamps were being introduced in these cities in the 1952, 1953, 1954 period. Notice just before that introduction in the cities in which they were coming along, prices had been averaging a little higher than in the other cities. With the introduction of stamps the price level came back down to the same level where it was before.

As a matter of fact, despite the adoption of stamps by food stores of the 15 cities, the average food price index for these cities stood in exactly the same relationship to the index for the 5 nonstamp cities in December of 1956 as it did in January of 1950.

The authors were certainly justified in concluding that the price trends which they found were

opposed to the hypothesis that stamps bring an upward divergence of food prices in the communities in which stamps are widely introduced into food stores.

Boston College study: Just this past August, Dr. Joseph D. O'Brien, associate professor of marketing at Boston College, published a study of food prices in Boston's supermarkets. His results support the Haring-Yoder study in that the chain having the lowest prices in the city of Boston over a period of 3 weeks gave trading stamps.

Gasoline prices: So far the evidence presented has referred directly to food. But similar evidence is also available for gasolineand this evidence is important to this committee, since the former is a large consumer of gasoline.

Exhibit III shows the prices asked by gasoline stations in Erie, Pa., on September 22, 1957.

Prices were gathered from each of the stations selling well-known brands and located on the right-hand side of Route 5 traveling from west to east through the city. Of the 9 stations, 7 gave stamps and 2 did not. Yet all sold gasoline for the same price, 27.9 cents per gallon.

Exhibit IV shows gasoline prices in a rural area. The stations covered were on the right-hand side of Route 20 and were located in the 36-mile stretch from Silver Creek, N. Y., to Depew, N. Y.

Prices were obtained from all the stations selling well-known brands so that no bias was introduced, as might have been true, had but certain stations been selected.

Again it is evident that low prices are not confined to those stations which avoid stamps. As a matter of fact, the lowest price was quoted by a stamp-giving station, while the two stations with the highest prices did not give stamps.

It has been said that if all of the retailers gave stamps that the consumer no longer has a choice. That simply is not true, sir. Only 12 percent of all retail sales in the United States are made currently by retailers giving stamps. It is higher in some fields, obviously.

By taking retail sales as a whole there is a vast choice open to the

consumer.

Stamps may reduce prices in nonstamp stores. Before leaving the influence of stamps on prices, I want to make one other point-one which is frequently overlooked. Taking a retail trading area as a whole, a good case can be made for the statement that stamps actually result in lower prices in that they force the nonstamp retailer to reduce his prices.

To be specific, when the Safeway organization-with a longtime policy of not giving stamps-found competitors taking business away from its units through an aggressive trading-stamp program, it retaliated by reducing its prices. And I have quoted here the executive of that company who gave that in a court record.

It is my observation that the A. & P.-another nonstamp retailerreacted in a similar manner. As a matter of fact, one study goes so far as to conclude that reducing prices is "the No. 1 weapon" of the nonstamp store.

It has been said that if all the retailers gave stamps that the consumer no longer has a choice. That simply is not true, sir; only 12 percent of all retail sales in the United States are made currently by retailers giving stamps. It is higher in some fields, obviously. By taking retail sales as a whole, there is a vast choice open to the con

sumer.

In support of this statement, it quotes the replies of owners or managers of 511 nonstamp food stores and 416 nonstamp drugstores as to the measures they adopted to combat trading-stamp competition512 percent of the food stores and 22.6 percent of the drugstores reported that they used lower prices.

The way the competitive system operates, I can call your attention to another illustration-I am referring specifically to what is going on as we are sitting here this morning-the introduction of stamps in the Chicago area. One of the major chains in that area happens to be Jewel Tea that has decided it will not adopt stamps, but it has already announced publicly that the way it is going to meet that competition is to reduce its own price. That is a natural reaction, the natural way a competitive system works.

Now, if these last observations are true, it may be that such studies as the Indianapolis survey by Professors Haring and Yoder actually understate the impact of stamps on prices. Because what such a study does is to relate prices in stamp stores to those in nonstamp stores.

It is a relative comparison.

If the prices of both stamp and nonstamp stores are reduced (which would be to the consumer's benefit), this general reduction in prices would not show up in a relative price comparison between stamp and nonstamp stores. While we need further investigation to substantiate this possibility, my observation of the way a competitive economy operates leads me to the tentative conclusion that the possibility is probably a fact.

Trading stamps and retail failures. Now I shift to another subject, and a new one for the committee, perhaps. One of the charges frequently levied against trading stamps is that they increase retail failures. Although this charge is stated in several ways, the general implication is that many retailers are forced out of business in any area in which trading stamps are introduced.

To subject this charge to a statistical test, I asked Dun & Bradstreet, Inc., to make some special tabulations of its retail-failure figures for 25 large cities. I asked them without any knowledge on their part as to what the figures were to be used for. They had no idea, but I simply asked them to make a special tabulation of their retail-failure figures for 25 large cities. They run these all of the time. They keep a steady count, as you all know, each month.

Specifically, the number of retail failures for (1) food stores, (2)' drugstores, and (3) dry-goods and general-merchandise stores were tabulated separately for each year from 1947 through 1956 for each of these cities.

The regular Dun & Bradstreet definition of "failure" was used in this study. While excluding discontinuances with outside obligations paid in full, this definition includes

those businesses that ceased operations following assignment or bankruptcy; ceased with loss to creditors after such actions as execution, foreclosure, or

attachment; voluntarily withdrew leaving unpaid obligations; were involved in court actions such as receivership, reorganization, or arrangement; or voluntarily compromised with creditors.

There are two exhibits which indicate the results of this particular study which I have here. On this exhibit, exhibit 6, perhaps, if you will look at that one first, I have placed the actual figures of the Dun & Bradstreet report. These are the 25 cities; here [indicating] are the cities with no stamp activity. Here are the cities with little or no stamp activity in these periods.

I have taken the exact figures for the 3 years, 1947, 1948, and 1949, as the period before stamps were important in these areas. I have taken 1954, 1955, and 1956 when stamps have been relatively important.

You will note, of course, what happened, that in this period there has been a great increase in retail failures. That is something we are all aware of, a great increase in food failures. From this number to this number during this period, of drug and dry-goods failures.

If you want to put those in percentage curves and have this period equal to 100, then for the current period the failure goes up to 243 or an increase of 243 percent, if I make the point clear. This being the base period, this being the other.

In all 25 cities there has been a great increase in failures. A significant thing, however, is to note that in the cities with no stamp activities, and in the cities where there was little activity back in the base period, but great activity in most recent years, in each case the failure rate the increases are below not only the cities without stamps, where that has not even been a factor, but, also, below the average for the 25 cities as a whole.

If I may make my point perhaps more clear, this will illustrate it: The height of this column represents the failures as compared with the 100 in the base period for the 25 cities. This represents the height of failures in the cities where there have been no stamp activities; in other words, in this case, the failures without stamps were not as great as the average in the 25 cities.

In these cases the failures were greater, but in the cities where stamps became important during this period, the increase is much less in each specific situation.

In this case, for example, the failure rate is 105 percent. Here, 144 percent; and here, 145 percent.

The whole point is not to justify a particular position or to explain what caused all of this.

Mr. COOLEY. May I interrupt?

Are you trying to make the point that the green stamps prevent bankruptcy?

Mr. PHILLIPS. No, sir.

Mr. COOLEY. That is your conclusion.

Mr. PHILLIPS. That is your interpretation.

Mr. COOLEY. That is the way I understood your explanation to us. Mr. PHILLIPS. No, sir. I am trying to

Mr. COOLEY. Where you have green stamps, you did not have any trouble with bankruptcy?

Mr. PHILLIPS. If I may continue with this one sentence: All I am trying to show is that there are many, many causes of retail failures. Mr. COOLEY. You know what this is all about.

The Republican administration goes along with this same stuff. We know that.

Mr. PHILLIPS. My only point is that retail failures are caused by dozens of faults, and to pick out a single one of them, trading stamps, and to say that that is the cause, I would not read the opposite into my figures here by any means.

Exhibit V makes it clear that the number of retail failures for each of the 3 retail fields increased substantially during the 1954-56 period as compared with the 3 years from 1947 to 1949.

Using the total 1947-49 failures as 100, in 1954-56 food failures rose to 243, drug failures to 238, and dry goods and general merchandise failures to 226.

However, for all 3 fields of retailing the increases in failure indexes were substantially less for those cities where the usage of stamps had gained the most as compared with both the 25 cities and with the cities without stamp activity (see exhibit VI).

To illustrate, whereas the 1954-56 failure index for drugstores in the 25 cities had advanced to 238 and 358 in the nonstamp cities, in the 9 cities showing a large gain in trading-stamp activity the failure index advanced just to 144.

In brief, there is nothing in these figures to suggest any validity to the statement that the widespread usage of trading stamps increases retail failures. As marketing students have long recognized, retail failures are a result of many factors, and to point to trading stamps as a major element is to distort the truth.

Finally, let us examine the position of the consumer in relation to trading stamps. The price studies cited earlier in this testimony point to the conclusion that by exercising her shopping ability the consumer may make her purchases in a stamp-giving store and still buy at prices comparable to those in nonstamp stores. In brief, she can get her stamps without extra payments.

So the question we now face is this: Are the stamps worth anything to her?

In answering this question we are on solid ground, since it has been studied by the Department of Agriculture.

Using manufacturers' list prices as a base, the Department computed the median merchandise value of a book of stamps at $3.74. They simply go into the store, find what that merchandise would cost, and then find out how many books it takes to get that merchandise, and they compute that value as $3.74, or 22 cents for each dollar spent to fill the stampbook.

Even when discount-house prices were used as the basis of comparison, the consumer obtained merchandise worth 1.55 to 2.29 percent of each dollar spent. So the answer is clear. Stamps do have value to the consumer.

As you have said, the consumer is a thrifty buyer.

Moreover, the Department of Agriculture adds that:

By and large, consumers appear to be satisfied with merchandise or gifts offered in exchange for stamps.

If still another study of the value of trading stamps is desired, it is available in exhibit VII. This is a study made by Prof. Eugene R. Beem. What he did was to find the value to the consumers.

97926-57--9