|
CONTENTS
Relative Impacts of Generic Advertising:
The Case of Salmon
Tables 1 & 2
Director’s Column
Next Meeting
|
NEC-63
Fall 2003
October 6-7, 2003
Embassy Suites Hotel
South Lake Tahoe, CA
Evaluation of
Non-Advertising Promotion Strategies
|
|
printable pdf file
Relative Impacts of Generic Advertising:
The Case of Salmon
Henry W. Kinnucan and Øystein Myrland
Auburn University and University of Tromsø
Studies of generic advertising often show benefit-cost ratios in excess
of 3:1. Yet producers wonder “If advertising is so profitable, why
am I losing money?” The answer is that generic advertising’s
effect on market demand typically is tiny. Benefit-cost ratios are large
because generic advertising expenditures in relation to product value
are small. Take beef as an example. The annual farm value of beef in the
United States is about $30 billion. The annual expenditure on generic
advertising for beef is about $25 million. Given these two numbers, how
much would demand have to increase to yield a 3:1 return? The answer:
approximately 0.25%. A sales increase of this magnitude is not apt to
be observable at the farm level. Nor is it likely to spell the difference
between profit and loss during times of economic stress. Thus, farmers
can be forgiven for being skeptical about generic advertising, particularly
when studies show large returns when farm prices are low. Still, the fact
remains that large returns are compatible with small effects. In fact,
in the beef example, if generic advertising were to increase demand by
5%, a not unrealistic expectation for a private firm that advertises,
the benefit-cost ratio would be on the order of 63:1!
The purpose of this article is to place generic advertising’s effects
in perspective. It does so by using Norway’s salmon promotion program
as a case study. This is a useful case study in that salmon prices are
determined in a world market, which means that results for salmon are
applicable to other traded goods such as cotton, soybeans, and citrus.
Also, Norwegian salmon is subject to tariffs, and a domestic feed quota
that restricts farm output. This provides an opportunity to assess the
interplay between government policies and generic advertising profitability.
MODEL
The model consists of 10 equations. The first four equations describe
the demand for Norwegian salmon in the domestic and international markets.
Specifically, export demand is segmented into three major customer markets:
European Union, Japan, and Rest-of-World. (The United States is excluded
due to a prohibitive tariff on Norwegian Salmon enacted in 1992.) Demand
in each market is assumed to be a function of price and advertising expenditures
by the Norwegian Seafood Export Council (NSEC). For the export demand
equations, NSEC advertising expenditures are interacted with the appropriate
exchange rate to take account of the fact that changes in the value of
the Norwegian currency affect the cost of promotion in foreign countries.
The next three equations link Norway’s wholesale price to wholesale
prices in the indicated export markets. These equations contain three
exogenous variables: the bilateral exchange rate, shipping costs, and
the export tax used to fund promotion. Exchange rates and shipping costs
permit evaluation of how currency realignments and changes in freight
rates affect demand compared to changes in advertising. The export tax
is included to take into account tax shifting. That is, when the export
tax is increased to provide monies for advertising, a part of that increase
is borne by foreign consumers. This “tax shifting” must be taken
into account when computing benefit-cost ratios, otherwise results will
be misleading.
The eighth equation in the model is a farm-wholesale price transmission
equation that links the farm market in Norway to the wholesale market.
This equation translates shifts in wholesale demand due to the advertising
to shifts in demand at the farm level. The ninth equation is a farm supply
equation that describes Norwegian producers response to price, and to
the feed quota. The feed quota has two effects: it shifts the supply curve
up, and it makes supply less price elastic. The latter effect has important
implications for advertising profitability, and for price stability. The
tenth equation closes the model by setting supply equal to demand. The
model was simulated using baseline data and parameter values for the period
1997-99.
RESULTS
Of key interest in this study is the effects of advertising
relative to other exogenous variables; namely, exchange rates, shipping
costs, tariffs, and the feed quota. To determine that, we simulated the
model for a 10% change in each of the variables. Representative results
are reported in table 1(for the complete set,
see Kinnucan and Myrland).
REFERENCE
Kinnucan, H. W. and Ø. Myrland. “Relative Impact of the Norway-EU
Salmon Agreement: A Midterm Assessment.” Journal of Agricultural
Economics. 53 (2002): forthcoming.
All variables have the expected effects. For example, an increase in
the export tax on Norwegian salmon into the EU reduces farm price and
production in Norway, increases the EU price, and reduces Norway’s
exports to the EU (table 1). Similarly, an increase
in advertising expenditures in the EU raises prices in both the EU and
Norway, and increases Norway’s production and exports. Strengthening
of the kroner against the euro makes Norwegian salmon more expensive in
the EU, which reduces Norway’s exports to that market, and lowers
Norway’s farm price and production. An increase in shipping costs
operates in a similar fashion as currency strengthening: EU price rises,
Norwegian price falls, as does domestic production and exports to the
EU market. Feed quota loosening lowers the farm price in Norway, expands
domestic production, increases exports to the EU market, and lowers the
EU price.
What is important for the purposes of this paper are the relative
impacts. Focusing first on farm price effects, the exchange rate rules.
Specifically, a 10% currency strengthening reduces farm price by 7.6%,
which is twice as large as the feed quota effect (3.7%), and 18 times
as large as the shipping-cost effect (0.41%). Advertising effects are
minute (0.17%), as expected since advertising expenditures in relation
to product value is low, about 1.0%. Still, the advertising effect (0.17%)
is larger than the tax effect (-0.061%), which means a simultaneous 10%
increase in the EU export tax and EU advertising expenditures would be
remunerative for Norwegian producers in that the net effect on farm price
is positive.
Turning to the EU price, a similar pattern obtains.
That is, the exchange-rate effect dominates (4.3%), followed by the feed-quota
effect (2.6%). Shipping costs (0.22%), advertising expenditures (0.12%),
and the export tax (0.03%) have effects that are unimportant by comparison.
Still, this does not mean that the latter variables are irrelevant from
a policy perspective. That depends on each variable’s influence,
but also on the magnitude of the variables’ changes. This is highlighted
in table 2, where we report model simulations
for actual changes in the exogenous variables as observed over the study
period.
Based on table 2, the most important variable
to affect prices, production, and trade flows between 1997 and 1999 is
the feed quota. This is because the change in the feed quota (12.4%) exceeds
the change in the exchange rate (- 4%). But notice that advertising has
a larger effect than the exchange rate. This is because advertising expenditures
increased 254%. Thus, although generic advertising’s ability to affect
the market is limited, with sufficiently large changes in expenditure,
its influence can be felt.
The alert reader may have noticed from table
2 that quota relaxation causes prices in the EU and Norway to drop
less than proportionately to the corresponding increases in quantity.
This is because the demand for salmon in the two markets is price elastic.
Thus, by loosening the quota Norway can enlarge its export earnings, and
increase the value of its farm production. This is in contradistinction
to many traditional farm products (e.g., cotton, milk, soybeans) where
supply restrictions raise farm value owing to an inelastic demand. But
the quota has one advantage for Norway in that it makes farm supply less
price elastic. This increases the returns to promotion, a perhaps unintended,
but nonetheless beneficial, effect.
CONCLUDING COMMENTS
This article’s basic theme is that generic advertising effects typically
are tiny. In the case of salmon, a 254% increase in generic advertising
expenditures in the European Union was shown to increase in the EU wholesale
price by a mere 3.0% and the Norwegian farm price by a mere 4.3%. Cause
and effect relationships of this magnitude are not uncommon in the commodity
promotion literature. This does not mean that generic advertising is unprofitable.
Indeed, our analysis suggests program intensification over the 1997-99
period yielded Norwegian producers a benefit-cost ratio of 3:1 (see Kinnucan
and Myrland for details). Rather, it indicates that generic advertising’s
ability to influence prices, production, and trade flows is limited.
One reason for the limited effect is that generic advertising expenditures
are small in relation to product value, as has been mentioned. A more
sophisticated reason comes from economic theory. In particular, industry
profit for a competitive industry is maximized when the following condition
holds:
 o
=  A
/ P
where o
is the ratio of advertising expenditures to product value in profit-maximizing
equilibrium,
A
is the advertising elasticity, and P
is the demand elasticity with respect to price in absolute value. Since
o
< 1, it follows that the demand elasticity sets the upper limit on
the advertising elasticity. For bulk agricultural commodities such as
cotton, milk, and soybeans where farm-level demand is highly price inelastic
( P
< 0.30 in the United States according to most estimates), it follows
that advertising response must be highly inelastic, on the order of A
< 0.02.
High-value products such as salmon, wines, nuts, fresh fruits, and raisons
that may have elastic demands at the farm level provide greater latitude
for advertising response. But even here we are not talking large effects.
For example, in the present study P
 1.2.
If one believes that o
< 0.05, i.e., the optimal advertising intensity for salmon is no higher
than 5%, then A
< 0.06. What this means is that to increase the demand for Norwegian
salmon by 6%, the industry would need to increase its advertising expenditure
by at least 100%. (We say “at least” because the demand shift
will cause price to rise, which will have a dampening effect on consumption.)
Thus, theory supports the intuitive idea expressed earlier that generic
advertising responses are minute. Their smallness explains why researchers
find advertising responses so difficult to estimate. It also explains
why producers may be skeptical about studies that show positive benefits.
But positive benefits and small effects are compatible, as the salmon
case shows.
|
