The paper presents an empirical analysis of the effect of a certified environmental label (the Nordic Swan), using a large Danish consumer panel with detailed information on actual purchases from the beginning of 19 97 to January 200 1 (weekly observations ). In the paper, we use a multinomial logit model to quantify the effect of the Swan label on consumers' choices between different brands of toilet paper, paper towels and cleaning products. In addition, the potential negative effect of the label scheme appears to be strictly related to the announcement of the label schemes (i.e. the launch phase).
12 Comparative Nordic surveys also showed that the meaning of the Swan label is significantly higher in Sweden and Norway (where the Swan label has also been used for a longer period of time).
Utility foundation for applying a model of brand choice to storable goods Let U denote ho usehold utility per unit of time, which is a ssumed to depend o n sub-utility
The multinomial model is based on the assumption that the utility value of each attribute is constant across all purchasing situations in the data period. From the beginning in the normalized utility equation (1a), the consumption decision is based on the expected cost per sub-utility unit V of replenishing inventories. In the purchasing situation, good qualities must be judged by their marginal use value in consumption.
The optimal quantity to buy in the given purchase situation is limited by the utility costs of transporting goods home, storing them, etc.
The statistical model
20Individual specific variables can also (or rather) interact with the characteristics of the choice. one can examine whether the swan label has different effects depending on income, age, education level, etc. Household specific variables do not differ across the alternatives and thus will not be identified as the individual variables in the same way as the brand attributes. Therefore, the omission of non-typical pack sizes seems unlikely to bias the estimated effects of the Swan label.
22 In the definition of trademarks, we did not distinguish between 'colour' (specially made for colored clothes) and 'ordinary'. which contain more bleach and are mainly used for white clothes) varieties of the same brand.
Data
Data sources
In these cases, the alternative specific constant is confused with false labels, so these labels do not contribute to the identification of the Swan effect. Thus, these two brands will make only a limited contribution to the empirical identification of the effect of the label on detergent choice. There has apparently been a large increase in the proportion of marked sightings, largely reflecting the increasing number of available brands that have acquired the Swan label.
The focus here is on brand choice rather than the number of packs or rolls sold in each choice situation.
Choices and variables in the model
The share of observations therefore does not necessarily correspond to the market share of the labeled goods, because each 'observation' describes the brand choice, and not the number of packages sold (nor the package size). For each consumer, the only price (P) captured in the data is that of the brand purchased. The test detailed results from what appeared to be a thorough washing test, as well as the environmental performance of the various compact brands.
31 Part of the marketing literature has paid attention to the empirical problems associated with evaluating the effect of coupons. Two of the biggest brands on the market (Arial and Omo) were highlighted as having. As a way to control for the effect of consumer report advertising, the dummy variable for good advertising (CGODPU B) has a value of 1 for Bluecare and T usindfryd), while another dummy variable (CBADPU B) has a value of 1 for Ariel and Omo. .
To control for the effects of the consumer report, alternative regressions will be run, including CTESTENV and CTEST (to control for the results of the consumer test) and CGODPU B and CBADPU B (focusing on the publicity effect of the consumer report), respectively. . Basically, we want to control all the factors that influence the sales of different brands of the analyzed products over time. Systematic information about the awareness of the different brands in these sales brochures was not available, but the awareness of brands in the sales brochures is most likely correlated with the sale variable31.
However, inclusion of various individual specific variables has little effect on the estimates of the parameters of the attribute variables. 34 As an additional way to check the effect of the consumer test, only estimates have been made on the choice of detergent.
Estimation re sults
- All brands included in choice set
- Result when choice set is conditioned on shop
- Interpretation of results: Price elasticity, WTP and model simulation
- Summ ary of results
This suggests that the estimated effect of the Nordic Ecolabel in this model is probably a lower limit for the effect of the environmental label. It is also worth noting that the variable for the effect of the consumer test report on paper towels was insignificant. For toilet paper and cleaning products, the effect of the Nordic Ecolabel is very similar in model 1 (e.g. 0.312 for toilet paper) and model 2 (0.382 for toilet paper).
Therefore, when looking at toilet paper in model 1 (which includes ONS) there is a positive sign on the SWAN dumm y coefficient in all 12 different stores. In one of the two cases, the estimated parameter for SWAN was also significant t. Finally, the estimated parameters rs for the indicators of the results of consumer reports confirm the results given in table 5.1.a.
The effect on the choice of detergents appears to be sensitive to the formulation of the model. An interpretation of the parameters on the price and label variables is provided below. Table 5.5 presents the averages of own price elasticities based on the estimated parameters presented in table 5.1.a, 5.2, 5.3 and 5.4.
In this light, the range of estimated WTP from 10.0% to 17.5% reveals a high consistency of the effect of the S wan label on toilet paper preferences. The third row shows the predicted proportion of labeled brands without the label effect.
Socioeconomic characteristics and effect of the Swan
Therefore, the effect of the asthma and allergy label in the study is controlled by the brand constants (ASC). It should also be noted that surveys conducted on the recognition and understanding of the Swan (see section 2.3) show that very few respondents mistakenly associated the label with healthier products. The opposite results obtained for the effect of income on the effect of the Swan seem to agree with results by Teisl et al.
Based on a hypothetical study of the choice of electricity supplier, the level of income did not significantly influence the effect of the brand (green electricity). In summary, it was found that the effect of the certified brand was lower for men and households with children in the choice of toilet paper and paper towels. In a number of cases, the effect of the brand also varied with income level and education, but the directions were not consistent across different product groups.
It is difficult to emphasize that the results from such studies should be interpreted with caution, since the laboratory in question may be related to other attributes of the labeled brands. However, paper towel brands bearing the Swan label are (on average) more expensive than unlabelled ones (suggesting they are of higher quality). This suggests that higher income households bought Swan labeled products because they were of higher quality, not because they carried the Swan la bel brand.
The proportion of labeled brands of paper towels and toilet paper tends to increase with the age of the household. The results in table 5.7 show, however, that the Nordic Ecolabel, after checking for brand quality, has not had a significant additional effect on the choice of brand.
Conclusion
Therefore, an important but unanswered question is whether the positive effect of the Swan ecolabel found in the Danish market will also be found in other countries, especially those outside the Nordic group. The potential effectiveness of providing information to consumers about the environmental characteristics of products as a regulatory tool. Cholesterol data and egg consumption in the United States: a non-normal and hetrosc edastic model with a double hurdle.
Title: Environmental labeling and consumer choice - An empirical analysis of the effect of the Nordic Swan. Explanation of the Swan label as proportion of respondents who recognize the label and proportion of the entire sample. For store 5, a large proportion of the purchase observations (47%) are not included for the same reasons.
For detergents, the elasticities come from the model with the 'publicity effects' of the consumer test (elasticities very similar in model with consumer test variables instead). Derived from parameters presented in table 5.1.a (all brands in choice set) 1) For d etergents of the model with the 'publicity effects' of the consumer test. In the definition of trademarks, no distinction was made between 'colour' (for colored clothes) and non-colour (for white cloth e) variations of the same bran.
11 Other without discount (Spar/Superbest/Iso etc). 1) Brands registered sold in the "trimmed" data set during 1997 to January 2001 (not necessarily sold over the entire time period). In a few cases, the number of observations used in the selected models conditioning on stores (Tables 5.2 and 5.3) is higher than indicated in Table A1 .3 (due to slight differences in data generation/trimming) .
