Norwegian Whitefish Export Demand Equation

The specification of equation (1) for the Norwegian whitefish export demand benefitted from the work of Kaiser (Reference Kaiser2015), Williams et al. (Reference Williams, Reimer, Dudensing, McCarl, Kaiser and Somers2016), and many of the export promotion studies they review. The dependent variable in equation (1) is the monthly demand for Norwegian whitefish exports (NWXD). The monthly price variable (NWPX) is the export unit value of Norwegian whitefish exports expressed in NOK per kilogram (kg). To account for changes in currency values against the NOK, NWPX was multiplied by a monthly trade-weighted exchange rate index (base period 2010) across the top ten importing countries in each month of each year. The trade weights are the monthly shares of Norwegian whitefish exports accounted for by the top ten importing countries from 2003 to 2017. The trade-weighted NOK exchange rate was developed using monthly trade weights calculated from monthly Norwegian whitefish export data (NSC 2018) and exchange rate data (IMF 2018). The exchange-rate-adjusted whitefish export price was then deflated by the FAO aquaculture fish price index (FAO 2018, Tveterås Reference Tveterås2018, Tveterås et al. Reference Tveterås, Asche, Bellemare, Smith, Guttormsen, Lem, Lien and Vannuccini2012) to account for the price pressure on Norwegian whitefish exports from growing global export supplies of aquaculture fish species.

To represent the purchasing power of consumers in importing countries in equation (1), the Gross Domestic Products (GDPs) of the top ten importing countries were converted to NOK, multiplied by their trade weights, deflated by an aggregate trade-weighted Consumer Price Index (base period 2010) for the same countries, and aggregated into a single real GDP measure for each month in each year (2010 NOK). The real aggregate GDP measure was then multiplied by the monthly trade-weighted exchange rate index to account for changes in currency values against the NOK.

We also tested for the significance of the Norwegian salmon price as a driver of whitefish export demand following Fousekis and Revell (Reference Fousekis and Revell2004) and Singh, Dey, and Surathkal (Reference Singh, Dey and Surathkal2012 and Reference Singh, Dey and Surathkal2014) among others. Like the whitefish export price, the Norwegian salmon export price was exchange-rate adjusted and deflated by the FAO aquaculture price index.

The goodwill variable G in equation (1) is a transformation of NSC whitefish export promotion expenditures to account for several key characteristics of the relationship between those expenditures and whitefish exports. First, NSC whitefish promotion expenditures were seasonally adjusted using the X13 technique (using EVIEWS 9.5 2018) to avoid confounding associated with seasonality. Second, the expenditures were exchange-rate-adjusted as done for the whitefish export price (NWPX) and inflation-adjusted using the trade-weighted Consumer Price Index (base 2010) in importing countries.

Third, a large body of literature supports the hypothesis that promotion expenditures have carryover or lagged effects (e.g., Nerlove and Waugh Reference Nerlove and Waugh1961, Lee and Brown Reference Lee and Brown1986, Ward and Dixon Reference Ward and Dixon1989, Williams, Shumway, and Love Reference Williams, Shumway and Love2002, Ghosh and Williams Reference Ghosh and Williams2016, Williams and Capps Reference Williams and Capps2017, Williams and Capps Reference Williams and Capps2018, among many others). Economic theory provides relatively little guidance as to the structure and length of this dynamic process, however. We follow a common practice in the literature of using the Almon polynomial distributed lag (PDL) formulation to account for the time lag in the impact of the NSC export promotion expenditures on Norwegian whitefish export demand.

The search for the pattern, polynomial degree, and period over which the promotion expenditures influence Norwegian whitefish export demand involved a series of nested OLS regressions. Conventionally, researchers use statistical criteria like the Akaike Information Criterion (AIC), the Schwarz Loss Criterion (SLC), or the Hannan-Quinn Criterion (HQC) to allow the data to suggest the optimal number of lags to include in the specification. Previous research on a broad range of food products suggests that the full impacts of promotion expenditures in a given month occur within no more than a year following the expenditure. Hence, we considered lags of NSC expenditures up to twelve months. The econometric results indicate the short-run and long-run or cumulative effects as well as the average length of time (in months) before changes in NSC export promotion expenditures begin affecting the Norwegian whitefish export demand.

Finally, NSC expenditures in equation (1) were transformed to capture the effects of diminishing marginal returns associated with advertising wearout (Kinnucan, Chang, and Venkateswaran Reference Kinnucan, Chang and Venkateswaran1993). To capture those effects, we implemented a logarithmic transformation of the exchange rate-adjusted, deflated, and seasonally adjusted NSC export promotion expenditures as is commonly done (Williams et al. Reference Williams, Reimer, Dudensing, McCarl, Kaiser and Somers2016).

While economic variables like purchasing power in importing countries, the export prices of whitefish, salmon, and other competing species, exchange rates, and promotion expenditures explain the long-term trends in exports, various events in global whitefish markets may have impacts in specific years or across many years. For example, the seasonal patterns evident in the volume of monthly Norwegian whitefish exports must be accounted for in the analysis. In addition, various qualitative events may affect global markets for whitefish in particular years and must be taken into account to isolate more accurately the effects of promotion on exports. The major global events with potential to impact Norwegian whitefish exports were identified in discussion with NSC personnel and treated as indicator variables in equation (1). Finally, we consider habit persistence or inertia using a one-period lag in the volume of whitefish exports.

Norwegian Whitefish Export Supply Price Response

Equation (2) of the model defines the response of Norwegian whitefish export supplies to changes in price. The critical link between supply response and the effectiveness of export promotion is illustrated in Figure 3. Assume, for example, that the promotion of the foreign demand for Norwegian whitefish shifts out the demand for whitefish exports in a given year from NWXDwo to NWXDw, as depicted in Figure 3 (where “wo” and “w” mean “without” and “with” promotion, respectively). Consistent with Chang and Kinnucan (Reference Chang and Kinnucan1991), the ad valorem levy on Norwegian whitefish exports behaves like an excise tax, the cost of which is shared by producers and consumers. Graphically, the effect of the levy is manifest with two export supply curves separated vertically by the ad valorem promotion levy (θ). The higher of the two curves represents the prices paid by foreign consumers at various levels of Norwegian whitefish exports (NWPX). For clarity of exposition and because our focus is the effects of promotion on Norway, only the prices received by exporters (NWP) given the levy are shown in Figure 3.

Figure 3.

Market and welfare effects of Norwegian whitefish export demand promotion

Given little or no export supply control as reflected in an elastic export supply like NWXS_e in Figure 3, most of the adjustment to a successful promotion program is manifest as an increase in export volume (${\rm NWX}_0^{{\rm wo}}$ to ${\rm NWX}_{\rm e}^{\rm w}$) rather than an increase in export price (${\rm NWP}_0^{{\rm wo}}$ to ${\rm NWP}_{\rm e}^{\rm w}$).Footnote ³ Even though the export price increase from the promotion-induced export demand shift is moderated by the vigorous export supply response in this case, export sales revenue is higher because the export quantity sold at the somewhat higher export price increases. The total cost of exporting also increases with the higher exports, however, so that the net benefit to the industry in terms of an increase in export surplusFootnote ⁴ (the horizontally lined area in Figure 3) is smaller than the net export revenue increase from the tax. The industry net benefit is positive unless the export supply is perfectly price elastic.

A less price-elastic export supply such as NWXS₀ in Figure 3, representing, perhaps, some effectiveness of supply controls, would result in a higher increase in the export price (${\rm NWP}_0^{{\rm wo}}$ to ${\rm NWP}_0^{\rm w}$) and a smaller increase in export sales (${\rm NWX}_0^{{\rm wo}}$ to ${\rm NWX}_0^{\rm w}$) than occurs with no (or ineffective) export controls. The resulting increase in the net benefit to the industry (the vertically lined area in Figure 3) is larger as a result.

Given absolute control of whitefish export supplies represented by a perfectly inelastic export supply (NWXS_i), the adjustment to a successful promotion program is manifest solely as an increase in the export price (${\rm NWP}_0^{{\rm wo}}$ to ${\rm NWP}_{\rm i}^{\rm w}$), leading to additional industry export revenue represented by the shaded rectangle in Figure 3. Thus, even if the export price increases considerably from the promotion, foreign sales cannot increase beyond ${\rm NWX}_0^{{\rm wo}}$ due to the supply restriction. Because export demand promotion has no effect on export volume, export suppliers’ costs of exporting do not change. Consequently, the increase in export revenue from the promotion is the addition to export surplus and the net benefit to the Norwegian whitefish industry.

Despite the generally accepted conclusion that fish supplies are exogenously determined, some level of long-run price responsiveness of the Norwegian whitefish export supply seems reasonable for various reasons. First, pushed by high prices and profits, overfishing established quotas has been a problem over the years (Gullestad et al. Reference Gullestad, Aglen, Bjordal, Blom, Johansen, Krog, Misund and Røttingen2014). Improved fishery management systems have reduced the incidence of overfishing, but overall compliance with the UN Code of Conduct for Responsible Fisheries (FAO 1995), which focuses on the prevention of overfishing, is less than perfect. Pitcher et al. (Reference Pitcher, Kalikoski, Short, Varkey and Pramod2009) rated Norway's fishery system the highest on compliance with the code at only about 60 percent.

Second, the Norwegian whitefish catch has been below the TAC in some years over the study period (HAVFISK 2015). During such periods, price likely plays at least some role in the corresponding supply changes. The demand-enhancing efforts of NSC during those periods helped push production towards the TAC limits by enhancing export demand and, thus, price.

Third, the export supply of whitefish is not the same thing as the domestic production of whitefish. Export supply curves are normally more elastic than their domestic counterparts because they are the difference between the domestic supply and demand curves. Although Norway exports about 95 percent of its whitefish supply (NSC 2018), the export supply can increase to some extent as the export price increases by drawing down the supply available to the small domestic market.

Finally, the quota levels themselves may be somewhat responsive to price trends over time. For example, after a period of increasing cod quotas and declining prices from 2009 through 2013, whitefish quotas were then reduced several years in a row with dramatic price supporting effects. Between 2013 and 2016, the export price of Norwegian frozen whole cod jumped by over 80 percent (Statista 2018). Then for 2017, with continuing high cod prices, Norway and Russia agreed on a total cod TAC of 890,000 mt, more than 11 percent above the ICES recommendation of 805,000 mt, leaving the TAC total little changed from the previous year. Then for 2018 when the ICES recommended a 20 percent cut in the cod TAC, Norway and Russia agreed on only a 13 percent cut to 775,000 mt. The higher negotiated levels of the TAC over the ICES recommendations in each year may reflect the higher prices for cod in export markets to some degree.

Studies using data before Norwegian fishery quota systems were fully binding found that the price elasticity of Norwegian fish supplies ranged from inelastic (Salvanes and Squires Reference Salvanes and Squires1995) to unitary elastic (Asche Reference Asche2009). In more recent modeling of world fish supply by the World Bank, capture fish supplies are assumed to be completely price inelastic (Msangi et al. Reference Msangi, Kobayashi, Batka, Vannuccini, Dey and Anderson2013). They assume the supply elasticity of aquaculture species, however, to be somewhere between 0.5 and 1.0. So over both capture and aquaculture, the supply of fish in the World Bank model would be somewhere between zero and 1.0. Asheim et al. (Reference Asheim, Dahl, Kumbhakar, Ogelend and Tveterås2011) reported a low price elasticity of Norwegian salmon supply in both the short run (0.091) and long run (0.141). Steen, Asche, and Salvanes (Reference Steen, Asche and Salvanes1997), Asche, Kumbhakar, and Tveterås (Reference Asche, Kumbhakar and Tveterås2007) and Andersen, Roll, and Tveterås (Reference Andersen, Roll and Tveterås2008), however, reported higher Norwegian salmon supply elasticities of about 1.4 to 1.5. The export supply elasticities estimated or assumed by studies of the NSC salmon promotion program have ranged from close to zero to 2.0 (Williams and Capps Reference Williams and Capps2017).

Allowing that the Norwegian whitefish export supply is not likely to be perfectly inelastic, an export supply elasticity estimate over 1.0 or even 0.5 seems equally unlikely given the nature of production controls. Nevertheless, we consider a range of alternative export supply elasticities from zero to two in our analysis to facilitate discussion of the possible extremes in the consequences of absolute to progressively less supply control and the implications for the Norwegian whitefish industry. In fact, we simulate five alternative levels of the whitefish price and export quantity effects of NSC whitefish promotion corresponding to five alternative Norwegian export supply elasticities: (1) 0, (2) 0.5, (3) 1.0, (4) 1.5, and (5) 2.0. The elasticities at the lower end represent absolute supply control (0) or at least reasonably effective supply control (0.5), while those at the upper end (1.5 and 2.0) represent the extreme limits of a failure to control supply effectively.

Counterfactual (Simulation) Analysis Procedure

To determine the effects of NSC promotion on Norwegian whitefish exports, two sets of scenarios were simulated with the model (equations [1] – [3]) using the econometric results for equation (1) from estimating the parameters for whitefish export demand: (1) a “with NSC export promotion” simulation (referred to as the “with” scenario) and (2) five “without NSC export promotion” simulations corresponding to the five alternative assumptions for the whitefish export supply elasticity in equation (2) of the model (referred to as the “without” scenarios). The use of simulation analysis in evaluations of the effectiveness of export promotion programs is standard practice in the literature (see Williams et al. Reference Williams, Reimer, Dudensing, McCarl, Kaiser and Somers2016.)

The with scenario represents actual history, that is, the export volume, value, and price of Norwegian whitefish which include the effects of the NSC whitefish export promotion expenditures over time. The without scenario analyses were conducted by setting both the historic values of NSC export promotion expenditures in equation (1) and the export levy (θ) in equation (2) to zero and then simulating the model to determine the volume, price, and value of Norwegian whitefish exports for each of the five alternative export supply elasticities used in equation (2) of the model representing the absence of NSC promotion over time in each case.

Differences in the simulated levels of Norwegian whitefish export volume, value, and price in the five with scenarios from those in the without scenario are direct measures of the effects of the NSC whitefish export promotion program in each case. No exogenous variable in the model (e.g., levels of inflation, exchange rates, income levels) other than the NSC promotion expenditures are allowed to change in any of the five without scenarios.

Benefit-Cost Calculation

A common measure of the “benefit” used in benefit-cost analyses of export promotion programs is the additional export revenue generated given that the promotion objective is to increase export sales. The simulated additions to Norwegian whitefish export revenue induced by NSC whitefish export promotion (RX) can be calculated as:

(4)$${\rm R}{\rm X}_{{\rm jt}} = {\rm NWP}_{{\rm jt}}^{\rm w} {\rm NWX}_{{\rm jt}}^{\rm w} -{\rm NWP}_{\rm t}^{{\rm wo}} {\rm NWX}_{\rm t}^{{\rm wo}} $$

where j represents alternative export supply elasticities and t refers to a particular year. Both the export prices (NWP) and quantities (NWX) in Equation (4) include the supply-shifting, excise-tax-like effects of the levy (θ) paid in part by Norwegian whitefish exporters (see Figure 3).

The Gross Revenue BCR (GBCR) for each export supply curve elasticity (j) considered given the excise tax effect of the levy is then calculated as the total additional Norwegian export revenue generated over the period of promotion from equation (4) per NOK of exogenous promotion expenditures (E) over that period:

(5)$${\rm GBC}{\rm R}_{\rm j} = \mathop \sum \limits_{{\rm t} = 1}^{\rm T} \displaystyle{{{\rm R}{\rm X}_{{\rm jt}}} \over {{\rm E}_{\rm t}}}$$

where T represents the last year of the promotion period. Note that E does not change as the supply elasticity changes in this ex-post analysis because we analyze the effects of a historically fixed and, therefore, exogenous level of promotion expenditures given different assumptions on the level of export supply responsiveness to price.Footnote ⁵

Because NSC promotion represents a cost of generating additional whitefish export revenue, the expenditures to promote each whitefish type in each year must be netted out of the additional export revenue for the respective whitefish type to arrive at the net export revenue BCR:

(6)$${\rm NBC}{\rm R}_{\rm j} = \mathop \sum \limits_{{\rm t} = 1}^{\rm T} \displaystyle{{{\rm R}{\rm X}_{{\rm jt}}-{\rm \; } {\rm E}_{\rm t}} \over {{\rm E}_{\rm t}}}.$$

Revenue-based BCRs are often the metric of interest among export promotion groups because revenue generation is typically an explicit goal of promotion. Even so, revenue-based export BCR measures may be misleading as promotion metrics because they fail to account for the additional costs required to generate the additional export revenue. To account for such costs, we can calculate the more appropriate change in export surplus (RS), as illustrated in Figure 3, to represent the “benefit” of promotion in year t:

(7)$$\eqalign{{\rm R}{\rm S}_{{\rm jt}} & = \lpar {\rm NWP}_{{\rm jt}}^{\rm w} -{\rm NWP}_{\rm t}^{{\rm wo}} \rpar {\rm NWX}_{\rm t}^{{\rm wo}} + {1 / 2}\lpar {\rm NWX}_{{\rm jt}}^{\rm w} -{\rm NWX}_{\rm t}^{{\rm wo}} \rpar \lpar {\rm NWP}_{{\rm jt}}^{\rm w} \cr & - {\rm NWP}_{\rm t}^{{\rm wo}} \rpar.}$$

The Net Export Surplus BCRs (SBCRs) are calculated by replacing RX_jt with RS_jt in equation (6). Although not precisely the same thing, export surplus, which approximates the Norwegian whitefish producer surplus,Footnote ⁶ can be considered a measure of the Norwegian whitefish industry export profit. The concept of the export surplus as profit is particularly applicable to the Norwegian whitefish industry with restrictive fishery quotas because a higher price in this case primarily increases the profit of individual fishers.

In Kinnucan and Gong (Reference Kinnucan and Gong2014), the change in producer surplus, a measure of the welfare effect associated with promotion, is approximated as a function of: (1) the own-price elasticity in the export market; (2) the export supply elasticity; (3) the export promotion elasticity; (4) the price elasticity of supply for domestic production; (5) the proportionate change in export price; (6) the proportionate change in export supply; and (7) the proportionate change in a “goodwill” stock variable in which the amount of advertising in the current period is a function of the current level of goodwill which in turn is a function of current and past advertising outlays. In our analysis, we explicitly account for all these variables in a structural model in which the parameters are econometrically estimated rather than assumed, except for the price elasticity of export supply. Consequently, our simulation analysis is tantamount, at least in theory, to that of Kinnucan and Gong (Reference Kinnucan and Gong2014).

Data

Monthly data on Norwegian whitefish export volume and value were provided by NSC along with data on their monthly export promotion expenditures for 2003 to 2017 (NSC 2018). Monthly data for exchange rates, inflation rates (CPI), gross domestic products for importing countries were all obtained from the IMF (2018). The FAO aquaculture price index developed by Tveterås et al. (Reference Tveterås, Asche, Bellemare, Smith, Guttormsen, Lem, Lien and Vannuccini2012) was obtained from Tveterås (Reference Tveterås2018). The indicator variables were developed based on information provided by the NSC.

Econometric Analysis of Whitefish Export Demand

For econometric estimation, equation (1) was operationalized as follows using monthly data from January 2003 through December 2017 (180 observations):

(8)$${\rm NWX}{\rm D}_{\rm t} = {\rm f\; } \bigg({\displaystyle{{{\rm NWP}{\rm X}_{\rm t}} \over {{\rm FAOP}{\rm I}_{\rm t}}}{\rm E}{\rm R}_{\rm t}\comma \; {\rm} \displaystyle{{{\rm NSP}{\rm X}_{\rm t}} \over {{\rm FAOP}{\rm I}_{\rm t}}}{\rm E}{\rm R}_{\rm t}\comma \; {\rm} \displaystyle{{{\rm MGD}{\rm P}_{\rm t}} \over {{\rm MCP}{\rm I}_{\rm t}}}{\rm E}{\rm R}_{\rm t}\comma \; {\rm PDL}\lpar {{\rm G}_{{\rm t}-{\rm n}}} \rpar \comma \; {\rm NWX}{\rm D}_{{\rm t}-1}{\rm SEA}{\rm S}_{\rm t}\comma \; {\rm} {\rm I}_{\rm t}} \bigg)$$

where NWXD = monthly Norwegian whitefish export volume; NWPX = nominal price of Norwegian whitefish exports paid by importing countries; NSPX = nominal Norwegian salmon export price; FAOPI = FAO aquaculture price index; ER = trade-weighted index of the NOK exchange rates of the top 10 countries that import whitefish from Norway; MGDP = trade-weighted nominal GDP of those same countries; MCPI = trade-weighted consumer price index of those same countries; PDL(G_t-n) = polynomial distributed lag model of the goodwill stock of Norwegian whitefish export promotion expenditures (G) over periods t, … t-n, where ${\rm G}_{{\rm t-n}} = \left({\displaystyle{{{\rm EX}{\rm P}_{{\rm t}-{\rm n}}} \over {{\rm MCP}{\rm I}_{{\rm t}-{\rm n}}}}{\rm E}{\rm R}_{{\rm t}-{\rm n}}} \right)$, EXP is the seasonally adjusted nominal NSC expenditures to promote whitefish exports, and n is the lag length determined as discussed previously; SEAS = seasonal monthly indicator variables with December as the reference month; I = is a matrix of indicator variables for various events affecting Norwegian whitefish exports; and t = time period. Equation (8) was estimated in double log form using ordinary least squares.

Model selection criteria (AIC, SIC, and HQC) were used to determine the most appropriate equation specification. The selected equation explains 83.2 percent (adjusted R2) of the variation in Norwegian whitefish export demand (Table 1). The signs and magnitudes of all estimated coefficients are consistent with prior expectations. As well, the within-sample mean absolute percent error (MAPE) over the sample period is 6.3 percent. The goodness-of-fit and the MAPE characteristics indicate a high degree of reliability of the econometric analysis.

Table 1.

Econometric estimation results for Norwegian whitefish export demand^a

Source: Computations by the authors using the software package EVIEWS 9.5 (2018).

^a All prices, GDP, and NSC whitefish promotion expenditures were exchange-rate adjusted. In addition, GDP and NSC expenditures were deflated by a trade weighted CPI representing the ten top importing countries.

Despite the relatively large number of explanatory variables, to avoid the inadvertent omission of any other important quantitative or qualitative factors, the residuals were examined to determine if a systematic pattern existed by testing for the presence of serial correlation. The Breusch-Godfrey LM test was implemented due to the presence of a lagged dependent variable (Breusch Reference Breusch1978, Godfrey Reference Godfrey1978). The null hypothesis of no serial correlation up to two lags was not rejected. This test is not only more general than the Durbin h-statistic but also more powerful statistically. In addition, ARCH tests, up to four lags of squared residuals, indicated the absence of any heteroscedasticity over time as well. The ARCH test regresses the squared residuals on lagged squared residuals and a constant. These tests further corroborate the appropriate specification of the model.

The issue of price endogeneity was examined using a Wu-Hausman test by comparing instrument variable (IV) estimates (two-stage least squares (2SLS) estimates) to OLS estimates (Wu Reference Wu1973, Hausman Reference Hausman1978). The hypothesis of the exogeneity of price was not rejected.Footnote ⁷ Like most export demand promotion evaluation studies, promotion expenditures are predetermined in our model (see Williams et al. Reference Williams, Reimer, Dudensing, McCarl, Kaiser and Somers2016). That is, the NSC determines the level of promotion expenditures based on decisions made by their board members. In turn, we explore the impact of these predetermined expenditure levels on the volume of Norwegian exports of whitefish.

The econometric results and associated statistics are provided in Table 1. The estimated whitefish export demand price elasticity of −0.31 is well within the range of those estimated by other whitefish demand studies. The estimated income elasticity of 0.25 is lower than reported by most of those studies except Hellandsjø (Reference Hellandsjø2015) for saithe.

Norwegian salmon and whitefish were found to be substitutes to some extent in importing countries (Table 1). While this finding is not new (see Singh, Dey, and Surathkal Reference Singh, Dey and Surathkal2012 and 2014, Fousekis and Revell Reference Fousekis and Revell2004), our estimated cross-price elasticity of 0.112 is comparatively low. Seasonality also was evident. Relative to December, Norwegian whitefish exports are highest in March (by 35.3 percent), February (by 29.2 percent), October (by 29.1 percent), September (by 25.2 percent), January (by 17.4 percent), November (by 14.1 percent), April (by 10.5 percent), and May (by 9.2 percent).

Selected qualitative events were also found to be drivers of Norwegian whitefish exports. Attention to these influential events is warranted in any economic analysis (Belsley, Kuh, and Welsch, Reference Belsley, Kuh and Welsch1980). While economic variables like purchasing power in importing countries, the export price of whitefish, exchange rates, and other variables largely explain the longer-term trends in Norwegian whitefish export demand, such events account for much of the deviation of export demand around the trend from year to year. To determine what events have impacted exports across time, we sequentially tested the effects of numerous events identified in discussion with NSC officials as potentially having had an impact on aggregate Norwegian whitefish export demand in various years. Of those events, the following were found to be clearly statistically significant over the study period, as indicated in Table 1: (1) weather-related events, particularly in 2006 and 2015; (2) the presence of bigger cod beginning in 2010; (3) increases in quotas for cod (2011 and 2013) and haddock (2009 to 2011 and in 2016); (4) the global financial meltdown in 2007 and 2008; and (5) the recession in 2008 and 2009. Moreover, statistical evidence of rigidities or inertia in Norwegian whitefish exports was found. The estimated coefficient of the lagged dependent variable (0.26) indicates that Norwegian whitefish export demand tends to respond quickly to changes in market forces.

The results exhibited in Table 1 also provide evidence that NSC expenditures have had a positive and statistically significant impact on Norwegian whitefish export demand over the study period. In the analysis, we considered polynomial distributed lags of order 2 and order 3 with and without endpoint restrictions for lag lengths of order 1 through 12. The specification that minimizes the model selection criteria consists of a second-degree polynomial with endpoint constraints together with a lag of two months. Thus, the impact of NSC export promotion is not felt all at once but instead is distributed over the current month of expenditure and the subsequent two months. Specifically, the econometric results indicate a short-run (contemporaneous) elasticity of NSC export promotion of 0.024 with a long-run (cumulative) elasticity of NSC export promotion of 0.079. Both results are in accord with those reported for many U.S. export promotion programs (Reimer et al. Reference Reimer, Williams, Dudensing and Kaiser2017, Williams et al. Reference Williams, Reimer, Dudensing, McCarl, Kaiser and Somers2016). The estimated long-run promotion elasticity is somewhat higher than reported for NSC salmon promotion which range from 0.0133 (Xie, Kinnucan, and Myrland Reference Xie, Kinnucan and Myrland2009) to 0.059 (Myrland and Kinnucan Reference Myrland, Kinnucan and Johnston2000). Kaiser (Reference Kaiser2015) and Xie (Reference Xie2015) report NSC salmon promotion elasticities of 0.036 and 0.054, respectively.Footnote ⁸

Simulation Analysis of NSC Whitefish Export Promotion

The simulation analysis examines the implications of various levels of export supply control from absolute to essentially no control on the effectiveness of NSC promotion on Norwegian whitefish exports and on the profitability of the Norwegian whitefish industry over the study period using the econometric results of estimating equation (8) in the model. The implications of absolute supply control for NSC export promotion were represented by a without scenario in which the export supply elasticity was set to zero in the model. The effects of progressively less effective export supply control were represented in four additional simulations in which the export supply elasticity was set at successively higher levels, including: (1) 0.5, (2) 1.0, (3) 1.5, and (4) 2.0 as discussed in the methodology section.

For each alternative export supply elasticity, the simulated differences in the levels of Norwegian whitefish export volume, price, value, and industry profit between the with and the respective without scenarios represent the additions to each as a direct result of investment by the Norwegian whitefish industry in the NSC promotion programs. Those differences are often referred to as the “lift” provided by a promotion program over the period of analysis. The “lift” achieved by a promotion program is the addition to total sales value or other industry measures as a result of the promotion, that is, how much higher sales or other industry measures were over time than they would have been if the promotion had not been conducted.

The results indicate that assuming absolute supply control (export supply elasticity of zero) over the 15-year period of the analysis, export promotion generates a substantial price lift of nearly 20 percent with no export volume response (Table 2). As the level of export control (unresponsiveness of the export supply to price) diminishes, the price lift also diminishes to as low as 2.3 percent for an export supply elasticity of 2.0, while the lift to export volume increases to 4.6 percent. At the same time, as the level of export control diminishes, the lift to export revenue and industry profit also declines.

Table 2.

Estimated lift^a from NSC whitefish export promotion under alternative levels of export supply control (export supply elasticities), 2003–2017

Source: Computations by the authors.

^a Addition to indicated measures from NSC promotion (from without promotion to with promotion).

^b Numbers in parentheses are the corresponding levels of export supply elasticities used in the simulations.

With absolute supply control over the 15 years of analysis, the export revenue and industry profit lifts from NSC whitefish export promotion would have reached a maximum of 16 percent and 32 percent, respectively, with no additions to actual sales of Norwegian whitefish (Table 2). With little or no supply control (export supply elasticity of 2.0), the export revenue lifts would have reached no more than 4.8 percent and 3.3 percent. Given the price inelasticity of the Norwegian export demand, the Norwegian whitefish industry maximizes the benefit from promotion when export supply control is most restrictive. These results may explain why Norwegian whitefish export prices and revenues tend to increase dramatically as supplies are constrained by quotas during the year.

Benefit-Cost Analysis of NSC Whitefish Export Promotion

Clearly, NSC export promotion has effectively boosted the price and revenues of Norwegian whitefish exports along with Norwegian whitefish industry profit regardless of the level of supply control. Perhaps more important to the industry is whether the export revenue and industry profit lifts induced by the promotion were substantial enough to more than cover their cost of financing the promotion. If not, then the promotion of whitefish should be discontinued because the promotion has cost stakeholders more than it has returned to them.

Because the level of benefits depends on the level of supply control, the returns per NOK spent on promotion also depends on the level of supply control. Based on equations (5) through (7’), the BCRs (GBCR, NBCR, and SBCR) for the NSC promotion of whitefish exports were calculated over the study period for all levels of export control (export supply elasticities) considered in the analysis. A BCR greater than 1.0 implies that the program has more than paid for itself. Otherwise, the program generates an economic loss because the benefit earned is less than the cost.

The GBCR and NBCR for the NSC whitefish export promotion program range from lows of 6.9 and 5.9, respectively, given little or no export control to 23.1 and 22.1, respectively, for absolute export supply control (Table 3). Thus, for every krone of NSC export promotion expenditure, the return to the Norwegian whitefish stakeholders in additional export revenue net of the promotion expenditures (NBCR) ranges between 5.9 and 22.1 kroner from little or no export control to absolute control, respectively. The corresponding SBCR, however, ranges from barely profitable (1.4) to highly profitable (22.1).Footnote ⁹ The SBCR is necessarily smaller than the NBCR because additional economic costs have been netted out of the additional export revenue. With no supply control, the NBCR and SBCR are equal, because without any lift in exports, there are no additional costs of exporting to subtract from the additional export revenue generated by the NSC promotion. The calculated NSC whitefish export promotion BCRs assuming less than absolute export supply control are in line with those calculated for salmon by Xie (Reference Xie2008 and Reference Xie2015), Kaiser (Reference Kaiser2015), and others who assume Norwegian salmon export supply elasticities of between 1.0 to 2.0. The results are also consistent with those found for cod and skrei by CAPIA AS (2016).

Table 3.

Benefit-cost ratios (BCRs) for NSC whitefish export promotion under alternative levels of supply control (export supply elasticities), 2003–2017

Source: Computations by the authors.

^a Numbers in parentheses are the corresponding levels of export supply elasticities used in the simulations.

^b Because this ex-post analysis considers the effects of a historically fixed (exogenous) level of promotion expenditures given different assumptions on the level of export supply responsiveness to price, the expenditures do not change as the supply elasticity changes.

^c Revenue-based export BCR measures may be misleading as promotion metrics because they fail to account for the additional costs required to generate the additional export revenue. The SBCR accounts for such costs and is, thus, the more appropriate promotion metric.

^d The domestic market accounts for only about 5 percent of the Norwegian whitefish catch so that the export surplus approximates producer surplus in this case.