Consumer Panel Data: A ZUMA research service

Papastefanou, Georgios Article Consumer Panel Data: A ZUMA research service Schmollers Jahrbuch – Zeitschrift für Wirtschaftsund Sozialwissenschaften. Journal of Applied Social Science Studies Provided in Cooperation with: Duncker & Humblot, Berlin Suggested Citation: Papastefanou, Georgios (2002) : Consumer Panel Data: A ZUMA research service, Schmollers Jahrbuch – Zeitschrift für Wirtschaftsund Sozialwissenschaften. Journal of Applied Social Science Studies, ISSN 1865-5742, Duncker & Humblot, Berlin, Vol. 122, Iss. 4, pp. 623-633, https://doi.org/10.3790/schm.122.4.623 This Version is available at: https://hdl.handle.net/10419/292031 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Schmollers Jahrbuch 122 (2002), 623-634 Duncker & Humblot, Berlin Consumer Panel Data: A ZUMA research service By Georgios Papastefanou 1. Introduction In modern societies research on private consumption has got a pervading relevance for marketing as well as for political administration and planning. Socio-economic transformations and challenges like sustainable growth or demand for innovations underline the necessity of empirically grounded knowledge on consumer behavior. On the other side consumer behavior patterns seem to be highly fragmented and diversified as social reality is transformed by lifestyle individualization and household structure differentiation. Social sciences have acknowledged this by theoretical perspectives (Friedrichs 1998). But nevertheless most lifestyle studies concentrate on consumer attitudes and valve orientations (Hartmann 1999). Only few studies combine behavioral data with data on household composition and as well as with data on attitudes. Actually, no standard micro data base for research on consumption behavior and its psychological and sociological correlates is available. On the other side, non-academic market research institutes like GfK in Nuremberg, collect detailed information and provide information services on consumer behavior (TWardawa, Wildner 1998). Special panel services run consumer panels producing continously longitudinal behavioral data on consumption of individuals and households. Aimed at supporting individual companies in marketing their specific products, consumer panels continuously are measuring the product flow into the household over years, what means that they are measuring the household's purchase event histories. These data are combined on a microdata level with socio-economic and demographic information on the household as well as with survey data on attitudes of specific reference persons in the household. So consumer panel data are of high potential for social science consumer analysis, but until now were not accessible generally. In 1997 ZUMA took the initiatve to establish an easy scientific access to consumer panel data. GfK, who runs a long established consumer panel, supported this initiative and gave the data to ZUMA for preparation and distribution to an interdisciplinary working group. After having proofed Schmollers Jahrbuch 122 (2002) 4 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.122.4.623 | Generated on 2023-04-04 12:30:56 624 Georgios Papastefanou the utility and manageability of the prepared consumer panel data in a conference (Papastefanou et al. 2001), GfK agreed in ZUMA's role in preparing and distributing the panel data as the ZUMA Consumer Panel Data of 1995. This paper describes the ZUMA Consumer Panel Data of 1995: (1) It gives an outline of the contents of the data. (2) It reports on selected topics using the potentiality of data. (3) It highlights the special feature of purchase history data and (4) it informs on the easy access to the micro data for scientific research and education. 2. Information in the data The ZUMA scientific use file of consumer panel data is based on the GfK ConsumerScan household panel. In this consumer panel GfK is running two independent samples with about 12000 households. They are designed as quota samples, quoted by federal state, administrative district, size of town and household size, with marginal distributions adapted to Mikrozensus distributions. The data, which are delivered by ZUMA as scientific use files are subsamples, containing those households, which in 1995 reported continuously their purchase diaries. These two subsamples comprise 9064 households. The ZUMA version file consists of two main types of data: a) data on each household's product purchases in 1995 and b) data on the socio-economic and demographic structure of the households at the beginning of 1994, 1995 and 1996 , as well as data on attitudes towards nutrition, environment and consumption of the person running the household in october 1994 and October 1995. These two types of data can be merged by a household identification number. Household purchases in 1995 Core of the consumer panel data collection are a household's purchases all over time with precise information on the timing of every purchase. In past, household budget diaries were used for measurement. Households used a booklet, in which they note all the purchases plus some of their characteristics.1 The ZUMA consumer panel data of 1995 are based on paperand-pencil household budet diaries. 1 Because of technological and product diversity the paper and pencil approach is replaced by home screening method of data collection. The participating households do have a manual screening device, by which they can collect the information of EAN code. The data then automatically are stored in a base station and are transfered later by telephone to the GfK data bank of the consumer panel. The home screening method was installed at GfK in 1998. Schmollers Jahrbuch 122 (2002) 4 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.122.4.623 | Generated on 2023-04-04 12:30:56 Consumer Panel Data: A ZUMA research service 625 In a methodological perspective one can characterize the data as event history data, which are not gathered retrospectively, but simultaneously with the process of product inflow into the household. The data are organized in asymmetrical data sets, in which the purchased product is unit of observation. So, for each household there may be an unequal number of purchase events. Their household assignment is achieved by a household identification number on every purchase record. The GfK ConsumerScan household panel is focused on fast moving consumer goods, especially on packed nutrition products and products of household convenience (see table 1). For each purchased product following variables are available: • Date of purchase • weekday of purchase • product category (measured in total purchases in 81 product categories, see table 1) • product subcategory (like flavor, taste etc.) • type of retailer • number of purchased products • type of price (standard vs. special price) • total quantity purchased • total amount of expenses • duration since last purchase • sort id of purchase (the sort id indicates the sequence number of every purchase in a household) • specific characteristics of the product, like type of package etc. Socio-economic and demografic situation of the household The socio-economic and demografic situation of the household is measured by a paper and pencil survey at the beginning of each year respectively. The following household characteristics are available in the ZUMA dataset: • federal state of residence • size of community • age of the household leading person • number of children (up to the age of 6, under the age of 14, under the age of 18) Schmollers Jahrbuch 122 (2002) 4 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.122.4.623 | Generated on 2023-04-04 12:30:56 626 Georgios Papastefanou Table 1 Product categories of the ZUMA consumer panel data file No of product category Sample 1 Sample 2 Label of product category No of purchased products in 1995 0 X X Window / carpet / toilet-cleaner 25.059 1 X X Only sample 1: tomato puree, only 6: pastasauces, both: Ketchup, spicesauces 47.651 2 X X Only sample 1: mayonese, tartar sauce, both: dressing 27.527 3 X X Special detergents 22.304 4 X X Detergents for dishes 34.556 5 X X Detergents for the household 20.940 6 X X Only 1: denture detergents both: toothbrushes/ mouthwash 26.818 7 X Tinned vegetables 63.275 8 X Milk 204.339 10 X X Water softener 29.542 11 X X Dental care products 55.497 12 X X Pure coffee (roasted) 143.194 13 X X Pure coffee-instand 31.576 14 X Salt 13.100 16 X Poultry 33.907 17 X X Frozen food 230.841 18 X X Tea 47.426 19 X X Cocoa 17.680 20 X X Spirituous beverages 59.129 21 X X General detergents 41.225 22 X Eats 233.124 23 X Mustard / horseradish 15.055 24 X Prefabricated cake 23.931 25 X Soap / washing lotion 13.105 26 X X Sparkling wine 30.127 27 X X Cider 3.229 28 X X Ground care 4.536 29 X X Bathing additives 33.206 30 X X Finished potato products 39.077 31 X Pudding / dessert 33.399 32 X Household tissues 12.265 33 X X Beer 131.245 34 X X Vermouth / appetizer 2.767 35 X X Wine / mulled wine 27.614 36 X X Alcoholfree uncarbonated beverages (fruit juice) 187.349 37 X Winned cabbage 48.502 38 X Crispbread 17.367 39 X Shoeand leather care 3.584 40 X X Salt for the dishwasher 4.703 42 X X Coffee stuff (like Caro) 3.564 43 X Delicatess salads 41.375 Schmollers Jahrbuch 122 (2002) 4 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.122.4.623 | Generated on 2023-04-04 12:30:56 Consumer Panel Data: A ZUMA research service 627 Table 1 (continue): No of product category Sample 1 Sample 2 Label of product category No of purchased products in 1995 44 X X Sherry / Port wine 3.343 45 X Pudding powder 14.454 46 X X Alcoholfree carbonated beverages (no mineral water) 155.447 47 X Cream cheese / soft cheese 144.046 48 X X Detergents for curtain 6.364 50 X Cream 74.009 51 X X Milk for coffee 111.247 52 X Air fragrants 2.683 53 X X Beverages with wine 5.292 54 X Gingerbread etc. 19.778 55 X Napkins 4.432 56 X Lights, matches etc. 2.730 57 X Vinegar 10.925 59 X Food care 1.113 64 X Sweet pastries 59.473 66 X X Animal food/cat litter (not prefabricated: only 6) 123.133 68 X Rusk 7.355 71 X Dry ready-to-serve meals 9.369 72 X Products for plants 3.598 73 X Curd/junkets 94.575 74 X Salty pastries 19.174 75 X Filter paper 12.648 76 X Instant soups 6.062 77 X Canned meals 11.244 78 X Yoghurt 119.270 79 X Flour 25.886 80 X Cleaning sponge 9.417 81 X Cheese 96.766 82 X X Detergents for the bath 6.278 83 X Baking powder/yeast 19.084 84 X X Mineral water 174.470 85 X Sandwich paper 18.756 86 X X Icecream 59.108 87 X Ready-to-bake blends 6.875 89 X Rice 16.158 90 X Cereals (Conflakes) 33.256 91 X Pasta 53.201 93 X Ready-made menues 4.641 94 X Roasted nuts etc. 58.008 99 X Toilet paper 3.245 Schmollers Jahrbuch 122 (2002) 4 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.122.4.623 | Generated on 2023-04-04 12:30:56 628 Georgios Papastefanou • occupational status of the household leading person • current occupation of the main income earner • occupation of the main income earner • former occupation of the main income earner • educational status of the main income earner • size of house • equipment of the household (washing machine, dishwasher, microwave, VCR) • garden ownership /-use • size of garden • pets (cat, dog, budgie) Socio-economic and demografic data in the ZUMA dataset are available for three waves of the ConsumerScan panel, namely of 1994, 1995 and 1996. The files of the different years can be merged by a household identification number, allowing for analysis of individual household structure changes. Attitudes of the person running the household Surveys on attitudes of specific persons in the household are part of the consumer panel. Under the assumption, that persons, who are running the household represent a kind of gate-keepers, they are the reference persons for the attitudinal survey. The ZUMA file version contains two attitude surveys of Octobre 1994 and Octobre 1995. They can be merged with the product purchase history data by a common household identification number. The attitude measures cover the following items: Attitudes towards nutrition: Items on medical health, natural, joy of discovering, german products, convenience orientation, slimness orientation, plain fare, full grain nutrition, superior savour, freshness orientation, pro branded goods, pro vitamines/minerals, uncritical style of nutrition Attitudes towards aspects of daily life: Items on tendency to innovate, traditional living, experience-orientation, nostalgia, mistrust towards new products, qualitiy-orientation, convenience-orientated cooking Attitudes towards environment: Items on ecological awareness, environmental behaviour, environment and mobility, state and industry Price consciousness of the household leading person: This variable contains a composite index which is based on ratings of three item pairs. Schmollers Jahrbuch 122 (2002) 4 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.122.4.623 | Generated on 2023-04-04 12:30:56 Consumer Panel Data: A ZUMA research service 629 3. Selected topics of analysis The ZUMA consumer panel data of 1995 offers unique possibilities for consumer behavior analysis. Since this type of data was not easily accessible for social and economic research in past there are only few empirical analyses. Wimmer (1995) for example used the data to estimate types of environmental oriented consumers to check for consistency of behavior and attitudes. By cluster analysis of the attitude data enclosed in the 1993 wave Wimmer (1995) specified two groups of households, those with a high environtalism score and those with low scores. Then he compared these two groups in relation to the bought amount of household cleaning and shower bath products. He could show that there was a consistency between favorability towards environtalism and actual purchasing of products, which presumable reduce environmental loading. As the data provided detailed information on specific product types, differentiated results were possible. Behavioral consistency was different for different product categories. High consistency was found in relation to amounts of softeners, sanitary cleaners and bubble bath ingredients. Low consistency was given in relato in to water closet cleaners, floor cleaners, universal detergents and shower baths. Similar results could be found for packing materials like nonreturnable bottles, glas and cans for carbonic acid and fruit beverages. A more recent analysis based on the ZUMA consumer panel data access is that of Thiele and Weiss (2001). The authors use the behavioral data to analyse empirically the determinants of variety in food consumption. Obviously no german analysis was done until now on this issue, the published research was based on US data expenditure survey data from 1972/73, 1977/78 and 1981, the last ones being restricted to urban households in the north-eastern region. Thiele and Weiss (2001) estimated several indexes for the household variety in food consumption for 149 food groups available in the purchase data and related them to several socio-economic household characteristics. They found that variety increases with household income, with the number of children aged between 6 and 18 years. Food diversity is also significantly higher if the family lives in larger cities in East-Germany and if the housekeeping person is not working full-time. Food product variety in single male households is significantly lower than that of female singles. A ZUMA symposium on the manageability and potential of the consumer panel data revealed a multitude of further relevant research issues. The results are reported in Papastefanou et al. (2001). Following, the topics of some of the contributions will be described shortly. In one paper, considering meanings of personal and social communication and exclusion by behavioral life style patterns, Ludtke and Schneider (2001) examine the data for life style strategies that can also be observed in Schmollers Jahrbuch 122 (2002) 4 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.122.4.623 | Generated on 2023-04-04 12:30:56 630 Georgios Papastefanou purchasing everyday consumer products. By including indicators of social position and attitudinal expression they further differentiate behavioral patterns along subjective and social meanings. One of the central findings is that social position and mentality correlates with fast moving consumer good purchases like food, beverages and hygiene products. Another analysis (Thogersen 2001) deals with the spread of sustainable behavior patterns of individuals. Thogersen asks if there is a mechanism of spillover, where a proenvironmental purchasing disposition (expressed in the purchasing of specific goods) is transferred to other environmentally relevant products. Using product information on packaging of purchased products he analyses also if spillover is supported by specific attitudes and values (Thogersen 2001). Ohr (2001) relates to the discussion on health risks in modern food production. He asks if purchasing of food products being related to health issues are influenced by attitudes towards nutrition and health. He finds consistency between purchasing and attitudes towards healthy nutrition, also if the involvement is on a medium level. Differentiating the expenses for healthy food products according to age, education and region he underlines that social position, mediated by socialization and cognitive competence, determine everyday life. Consumer panel data can be used also for critical evaluation of theoretical economic models on price formation and examination of dynamics and variation of retail prices. Fengler and Winter (2001) analyzed the purchases of coffee, as it is a product being bought frequently and having changing prices. Their results show the importance of psychological price thresholds for price settings and changes along changes in costs. The event history information potential of consumer panel data is utilized in Fachinger (2001), who is interested in risk evaluation by consumers. He argues that individual risk evaluation should be best detected when considering continuity of purchasing behavior. The availabe information on the timing of a purchase act allows to apply hazard rate models on purchases of coffee, detergents and yeast. Information on changes in net household income between two waves is used in the analysis of Papastefanou (2001). He examines the hypothesis that consumption of alcoholic beverages (with more than 20 Vol. alcohol) is correlated with household income position. The panel data made it possible to separate income change from income level effects on the bought amount of liquors. By this, empirical support could be found for a positive income level effect, but a negative income change effect. Schmollers Jahrbuch 122 (2002) 4 OPEN ACCESS | Licensed under CC BY 4.0 | https://creativecommons.org/about/cclicenses/ DOI https://doi.org/10.3790/schm.122.4.623 | Generated on 2023-04-04 12:30:56