A Tobit . .\Nalysis of Fac:Ors Affecting Vegetable Expenditure
Total Page:16
File Type:pdf, Size:1020Kb
A Tobit ..\nalysis of Fac:ors Affecting Vegetable Expenditure ?atter~s in U. S. Households by John McDonough Love Thesis submitted to the Faculty of the Virginia Polytechnic Institute and State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE in Agricultural Economics APPROVED: Oral Capps, J.x., Chairman Charles W. Coale, JT. Randall A. Kramer August 1982 Blacksburg, Virginia Acknowledgements The author expresses his ::;incere appreciation to Dr. Oral Capps, Jr., committee chairman, for the grace and generosity of his guidance while completing this thesis. He is also indebted to his other com- mittee members, Drs. Charles W. Coale, Jr. and Randall A. Kramer for their constructive criticisms in preparation of this manuscript. For technical assistance, he is indebted to the labors of Tom Finn and John Harvey. The author would like to gratefully acknowledge the financial support provided by S-165 Regional Project on U.S. Food Demand and Consumption Behavior, under Virginia Tech Agreement Number 230-01- 023-001-6224270. ii Table of Contents Acknowledgements. ii List of Tables. v List of Figures . vii Chapter I: Introduction .•. 1 Background . • . 2 Literature Review. 9 Problem Statement. 19 Objectives •••. 21 Scope ...... 21 Thesis Organization .. 22 Chapter II: Theoretical and Statistical Models 23 Introduction . 24 Partitioning Vegetables into Weakly Separable Groups 24 Economic Model • • . • • • . 28 Statistical Models • 31 Tobit Analysis . • • 35 Chapter III: Data Description and :iethods of Analysis. 39 Introduction • . • . • 40 Income . • • . • . • . 41 Age-Sex Composition. 42 Education •.• · ..•. • 44 Race of the Household Head . • 45 Fqod Stamp Participation ·: . 46 Earner Composition , ; -. • 47 Population Density of Residence. • 48 United States Region • . • • • • • 50 Methods of .<\nalys:!s. • . • . • • . 51 Chapter IV: Results and Conclusions. 55 Total Vegetable Expenditure ...• • 56 Fresh' and·-Canned Vegetable Expenditure . 62 Vegetable Subgroup Expenditure . • • • • • 72 Household Vegetable Expenditure Profile .• 90 iii Chapter V: Summary, Conclusions, and Implications for Further Research. • 93 References. 98 Appendix A. .101 Appendix B. .114 Vita. .172 Abstract iv List of Tables Table 1. Two vegetable subgroups based on frequency of preparation . • 26 2. Five vegetable subgroups and three processed forms which are used to estimate the impact of socioeconomic and . demographic factors on household eA"Penditures • • • • • • • 29 3. Variable notation and description for statistical model •• 34 4. Description of Tobit model and some results from McDonald and Moffitt (1980) • • • • • • • • • • • • • 36 5. A comparison of proportions of nonpurchasing households between the BLS CEDS, second sample year, and its subsample of 2,000 households randomly selected •• 52 6. Sutmnary statistics for Tobit analysis of household expenditures on total vegetables. • • • • • • • • ••• 57 7. Calculated elasticities.from.Tobit coefficients for total vegetables ••••••••••• . • • . 61 8. Sutmnary statistics for Tobit analysis of household expenditures on ·canned vegetables • • • • • • • • • • • 63 9. Calculated eJ.asticities from Tobit coefficients for canned vegetables • • • • • • . 65 10. Sunnnary statistics for Tobit analysis of household .. expenditures on fresh vegetables. • • • • • • • • ••• 66 11. Calculated elasticities from Tobit coefficients for fresh vegetables •••• • • • • 68 12. Summary statistics for Tobit aJ1alysis of household expenditures on dark green and deep yellow vegetables • • • 74 13. Calculated elasticities from Tobit coefficients for dark green and deep yellow vegetables • • • • • • • • • 76 14. Summary statistics for Tobit analysis of household expenditures on light green vegetables. • • • • • ••• 77 15. Calculated elasticities form Tobit coefficients for light green vegetables. • • • • 79 16. Sunnnary statistics for Tobit analysis of household expenditures on tomatoes. • • • • • • • • • • • • 80 v Table Page 17. Calculated elasticities from Tobit coefficients for tomatoes •••••••• • • • • 82 18. Swmnary statistics for Tobit analysis of household expenditures on potatoes. • • • • • • • • • • • • ••• 83 19. Calculated elasticities from Tobit coefficients for potatoes •.••••••• . .. • • • • 85 20. Summary statistics for Tobit analysis of household expenditures on other vegetables. • • • • • • • • ••• 86 21. Calculated elasticities from Tobit coefficients for other vegetables •• . • • • • • 88 22. Profile of two-week expenditures on vegetables by households under selected conditions of annual income and age-sex composition • • • • .•.•.• , • • • • • • • • • • • 91 vi List of Figures Figure 1. Changes in per capita consumption of vegetables, since 1967, in fresh, frozen, and canned forms •• 4 2. Changes in total per capita consumption of vegetables and percent total per capita consumption in fresh, frozen, and canned forms, since 1967 .•••••••• . 5 vii Chapter One Introduction 1 2 3acl.<ground In American households, vegetables significantly contribute to meeting the requirements for a balanced diet. While only occupying 5 to 6 percent of the average total household food budget, per capita consumption of vegetables is ap~roximately 20 percent of total per capita food consumed in the United States. An interest in maintaining or improving the diet of a particular sector of society naturalt"y leads to questions about current vegetable consumption and expenditure patterns. However, a scarcity of current research exists to answer questions about how socioeconomic and demographic factors affect the propensity to purchase vegetables and the expected value of vegetable expenditures. For obvious economic reasons, 'legetable producers and processors arc interested in factors affecting household demand for various vegetables and their associated product forms. Also, marketers and retailers arc keenly aware of the rising value of supermarket shelf space needed to adequately satisfy the demand for quality and variety in vegetable supply. In general, policymakers, producers, and marketers of food need to 1.<now as much as possi~le about how socioeconomic and demographic variables affect household food demand. Therefore, the present study focuses upon expenditures for vegetables made by U.S. households over a two-week period during the years 1972 to 1974. The aim of this research is to gain a better understanding of how socioeconomic and demographic factors affect household expenditure patterns for vegetables. 3 Vegetables arc important in a balanced Jict, as they arc rich sources of fiber, carbohydr.gtcs, minerals and vitamins. For example, the dark green and deep yellow vegetables (broccoli, collards, carrots, etc.) arc high in vitamin A content and potassium, and potatoes and corn arc high in carbohydrate content. Also, onions, peas, and beans arc all high in thiamin content (USDA, 1975). The recently formed guidelines for an improved quality of American diets recommended increased consumption of vegetables (U.S. Congress). Since the middle 1960s, per capita consumption of vegetables has increased gradually (USDA, 1977). The sharpest increase occurred in consumption of frozen and canned forms, while there was virtually no increase in per capita consumption of fresh forms (Figure 1). The United States Department of Agriculture estimates that, tn 1980, U.S. per capita consumption of vegetables was 207 pounds, 70 percent fresh, 25 percent canned, and 5 percent frozen (Figure 2). Since 1960, per capita consumption of the dark green and deep yellow vegetables has undergone minor proportional changes (Appendix Figures A.1-A.3). As fresh, the proportion made up of broccoli has increased, while the share going to carrots has decreased. As frozen, the proportions going to broccoli, brusscl sprouts, and carrots have increased. As canned, the proportion of dark green and deep yellow vegetables composed of carrots has increased slightly. Similarly, among the light green and "other" subgroups of vegetables (Figures A.4-A.9), per capita consumption 4 130 120 110 Cmned - 100 Q ~ ... 90 "'~ ..,," "'c - 70 75 80 110 100 Fresh WI _, 1967 68 69 70 71 , - ;3 74 7§ 76 77 ;8 79 80 al 82 Year Figure 1. Changes in per capita consumption of vegetables, since 1967, in fresh, frozen, and canned forms. (Source: USDA, Food Prices, Consumption, and Expenditures) 5 llO, ~ ,...I ...: ....,. lOO ,._.. - -- oo- ..:: ..... 90 o~ 100 Frozen 90· 80 c.nned 70 ....... ... 0= 60 u.. -.. .. ...l! .. ;; .. = 50. ~8 r--0 " .....&I 40' F~aah :: .. .."' .. .. .. 30 .."> " 20 10 1967 68 69 70 7l 72 73 74 75 76 11 78 79 80 81 s: Year Figure 2. Changes in total per capita consumption of vegetables (upper) and percent total per capita consumption in fresh, frozen, and canned forms (lower), since 1967. (Source: USDA, Food Prices, Consumption, and Expenditures) 6 of vegetables is characterized by minor long-term changes in the proportional mix of vegetables within a given processed form. Focusing upon tomatoes and potatoes, tt is apparent that long- term changes arc minor in per capita consumption of various processed forms. The proportion of per capita consumption of tomatoes going to catsup, chili sauce and paste has increased, while that going to juices has decreased (Figure A.10). The proportions in whole and pulp/puree remained stable. The mix of product forms of potatoes has shown a gradual shift away from fresh to frozen, while dehydrated, chipped and stringed, and canned potatoes remained stable (Figure A.11). In general,