Volume 49, Issue 1 Factors Affectng Frequency of Fast Food Consumpton Sayed Saghaan a! and Hosen Mohammad b a Professor, Department of Agrcultural Economcs, Unversty of Kentucky, 314 Barnhart Buldng, Lexngton, KY, 40546-0276, USA b Assocate Professor, Department of Agrcultural Economcs, Ferdows Unversty of Mashhad, Mashhad, Iran Abstract Progress n scence, technology, and ndustry has changed human lfestyle and especally food consumpton habts. For many famles, especally n large urban centers, fast food consumpton has become routne. To determne factors affectng the frequency of fast food consumpton, ths study surveyed 396 famles usng Posson and negatve bnomal regresson models. The man reasons that households consume fast food are a shortage of tme and for entertanment. Polcy makers could use socal marketng tools to control the growng trend of fast food consumpton. Keywords: consumpton, fast food, health, marketng tools, nutrton! Correspondng author: Tel: (859) 257-2356 Emal: ssaghaan@uky.edu March 2018 22 Volume 49, Issue 1
Introducton All socetes, especally n urban areas, face changes n transportaton, communcaton, nutrton, and health that affect lfestyle and eatng habts. Snce the 1950s when fast food entered markets hamburgers, pzza, french fres, and other foods that can be prepared and served quckly at relatvely low prces have become more and more popular (Song, 2016). Fast food consumpton has the potental to be harmful to human health. Afolab, et al. (2013) showed that fast foods are concentrated sources of energy, low n fber and hgh n detary cholesterol, and could sgnfcantly contrbute to detary cholesterol ntake, wth mplcatons for cardovascular health. In addton, the amount of salt n fast food ncreases the rsk of heart attacks, and consumpton of fast food and hgh-calore condments such as carbonated beverages and sugary sauces double the effects of such foods. A report by the World Health Organzaton (WHO) s Internatonal Agency for Research on Cancer (WHO, 2015) ndcated that there was enough evdence to consder processed meats to be group 1 carcnogens because of a causal lnk wth colon cancer. The IARC s experts concluded that each 50 g (1.8 oz) porton of processed meat eaten daly ncreased the rsk of colorectal cancer by 18%. The labels people attach to fast food are always hgh n calores, low n nutrtonal value, obesty, and addtves (Song, 2016). Despte wdespread knowledge about the harmful effects of fast food on human health, the demand for fast food restaurants s stll growng. Many studes have nvestgated factors that affect people s tendency to consume fast food (Park, 2004; Sahagun and Vasquez-Parraga, 2014, Song, 2016). Gven the mportant role of fast food consumpton on obesty, cardovascular problems, and other related dseases, ths study attempts to nvestgate the man factors affectng the frequency of fast food consumpton by households n a month n the cty of Mashhad n Iran. Lterature Revew Even though some consumers are aware of the consequences of consumng fast food, fast food consumpton s ncreasng around the world (Xue et al., 2016). Hearst et al. (2013) showed that frequent consumpton of fast food menu tems hgh n fat, sugar, and sodum contrbute to poor detary qualty, ncreasng ndvduals rsk for det-related chronc dseases. Consumpton of fast food has been assocated wth food safety problems n some developng countres (Omar and Frempong, 2016). Socoeconomc and demographc varables such as age, educaton, ncome, hours spent at work, and the number of household members have sgnfcant effects on the probablty of consumng fast food (Fannng, Marsh, and Stegert, 2002). Fast food s relatvely cheap, less tme consumng, and tasty, and these advantages ncentvze consumers to ncrease consumpton. Austn et al. (2005) showed that fast food sales ncreased 900% between 1975 and the md- 2000s: from $16 bllon to $153 bllon. Advertsng and brandng are powerful tools that affect food choces (Chrstan and Gereff, 2010); chldren are often targeted by fast food marketng March 2018 23 Volume 49, Issue 1
messages. Lee and Len (2015) concluded that fast food advertsng and other marketng tools postvely nfluence fast food consumpton and that fast food advertsers make profts at the expense of chldren s health. Socal marketng the applcaton of marketng prncples, tools, and technques to nfluence socally desrable behavors n a target audence s one potental soluton for controllng the growng consumpton of fast food (Lee and Kotler, 2015). Carns and Rundle-Thele (2014) showed that socal marketng can promote healthy eatng and the effectveness of socal marketng on healthy eatng could be enhanced. Few studes have addressed consumers preferences for fast food consumpton n developng countres such as Iran. Ths research analyzes factors affectng the frequency of fast-food consumpton of products lke hamburger and sausage n Mashhad, the captal of Iran s Khorasan Razav provnce. Mashhad s the second largest cty n Iran, wth a populaton of around 3.2 mllon n 2016. We hypothesze that socoeconomc and marketng factors contrbute to the frequency of fast food consumpton. Ths study nvestgates the consumpton frequency of fast food among Mashhad households over the course of a month usng Posson and negatve bnomal regresson models and consders the effects of marketng tools on consumpton frequency. Methodology Ths research nvestgates the factors affectng frequency of monthly fast food consumpton among Mashhad households. Hence, the dependent varable s a count varable showng the number of tmes each household purchases and consumes fast food. The use of lnear regresson models for count outcomes can result n neffcent, nconsstent, and based estmates (Long and Freese, 2001), but there are models that deal explctly wth propertes of count outcomes such as the Posson regresson model. Wth ths model, a Posson dstrbuton determnes the probablty of a count, and the mean of the dstrbuton s a functon of the explanatory varables. The condtonal mean of the outcome s equal to the condtonal varance. If condtonal varance exceeds the condtonal mean, then the negatve bnomal regresson (NBR) model s used (Long, 1997). Ths method can be used when the objectve s the descrpton of a count varable wth some explanatory varables. After some assumptons and calculatons, the NBR model can be estmated by maxmum lkelhood (ML) method. The lkelhood equaton s N N 1 1 Γ ( y ) 1 + α α α µ y (1) L( β y, X ) = Pr( y x ) =.( ) ( ), 1 1 1 y! Γ ( α ) α + µ α + µ = 1 = 1 1 where µ = exp( x β), υ = α for α > 0, and υ s the mean of the gamma ( Γ ) dstrbuton for δ. After takng logs, the log-lkelhood equaton can be maxmzed usng numercal methods (Long, 1997). The NBR model s one of a class of models constructed by mxng the Posson dstrbuton wth a second dstrbuton ( Γ ), whle other dstrbutons and mxtures can also be March 2018 24 Volume 49, Issue 1
used. Fnally, we use factor change to nterpret the results of the Posson regresson model (PRM) or the NBR model. The data used n ths study were collected through feld study n 2016. The sample sze was determned and stratfed random samplng was used to sample 13 regons of Mashhad, based on the classfcaton of urban areas by the Muncpalty of Mashhad. Results The dependent varable s the frequency of fast food consumpton by households n a month. The explanatory varables and ther descrpton are lsted n Table 1. Table 1. Model Varables and Descrptons Varable Descrpton Gender of household head Women =0, men=1 Educaton level Income Composton awareness Spouse occupaton Other meat consumpton Avalablty Qualty Age Advertsng Informatve label Prce Years educaton of household head Household s monthly ncome Awareness about composton, nutrtonal value, producton methods, and ngredents of fast food (aware=1, not aware=0) Employed =1, not employed=0 Amount of other meat consumpton n a month Effect of avalablty on consumpton (low=0, hgh=1) Effects of qualty on consumpton (low=0, hgh=1) Age of respondent Effects of advertsng (low=0, hgh=1) Effect of label on consumpton (low=0, hgh=1) Effect of prce on consumpton (low=0, hgh=1) Table 2 presents the man reasons and prortes that households gave for consumng fast food products (sausages, salam, and hamburgers). Entertanment was lsted as the man reason (frst and second prorty) for 52% of respondents. Shortage of tme was lsted as the man reason by 50% of respondents. Hence, the most common cause of fast food consumpton such as sausages and hamburgers by Mashhad households are entertanment and shortage of tme. The results of Posson regresson and NBR models are shown n Table 3. Comparng the goodness-of-ft measures, the NBR model performs better than the Posson model. Consderng that the varance of consumpton frequency s larger than ts average, the NBR model s selected and hence, only the coeffcents of the NBR model are nterpreted. March 2018 25 Volume 49, Issue 1
Table 2. Man Reason for Fast Food Consumpton by Households Man Reason Prorty 1 2 3 4 Other Shortage of tme Frequency 147 37% 52 13% 34 9% 39 10% 124 31% Taste and flavor Frequency 77 19% 78 20% 65 16% 46 12% 130 33% Habt Frequency 6 1.5% 45 11.3% 60 15.1% 59 15% 226 57% Chld penchant Frequency 30 8% 36 9% 50 12.6% 47 11.9% 233 59% Entertanment Frequency 131 33% 75 19% 55 14% 25 6% 110 28% Table 3. Man Factors Affectng Fast Food Consumpton Frequency Posson Regresson Model NBR Model Standard Standard Varable Coeffcent Error Z stat. Coeffcent Error Z stat. Constant 1.083 ** 0.54 2 0.984 * 0.595 1.65 Gender 0.053 0.104 0.51 0.066 0.1 0.66 Educaton -0.041* 0.024-1.7-0.052 *** 0.021-2.47 Income -0.151 ** 0.073-2.07-0.141 * 0.077-1.82 Spouse occupaton 0.233 ** 0.119 1.95 0.264 ** 0.119 2.2 Other meat -0.191 *** 0.076-2.5-0.21 *** 0.082-2.6 Age -0.019 0.034-0.58-0.016 0.034-0.48 Composton awareness -0.0065 ** 0.003-2.15-0.006 ** 0.003-2.03 Qualty ndex 0.0004 0.003 0.12 0.0001 0.037 0.05 Avalablty 0.075 0.051 1.49 0.084* 0.050 1.68 Advertsng 0.045* 0.026 1.73 0.043* 0.024 1.79 Label -0.12 0.17-0.7-0.14 0.18-0.77 Prce -0.053 0.046-1.16-0.048 0.046-1.05 Alpha stat. 0.5757 *** 0.06 9.4 Person stat -1,351.8 *** Devance stat 1,318.3 *** -2log lkelhood 2,435.8 2,002.76 McFadden R 2 0.031 0.011 Cragg Uhler R 2 0.18 0.054 AIC 6.24 5.15 BIC 157.8-569.9 Notes: Sngle, double, and trple astersks (*, **, ***) ndcate sgnfcance at the 10%, 5%, and 1% level, respectvely March 2018 26 Volume 49, Issue 1
Accordng to Table 3, the educaton level of household head, ncome, spouse occupaton, other meat consumpton, awareness about composton of fast food, avalablty of fast food, and advertsng have sgnfcant effects on the amount of fast food consumed by households. The results show that for a unt change n ncome, the expected count of fast food consumpton decreases by a factor of exp (-0.141), or 0.87 unts, holdng other varables constant. Hgher educaton levels, other meat consumpton, awareness about the composton of fast food, and ncome all have a negatve and sgnfcant effect on the frequency of fast food consumpton, whle spouse occupaton and avalablty of fast foods have a postve effect on the frequency of fast food consumpton. Fnally, by advertsng fast food, the expected frequency of fast food consumpton ncreases by 4% compared to the case where there s no advertsng, holdng other varables constant. Moreover, the results n Table 3 show that varables such as prce do not have any sgnfcant effect on the frequency of fast food consumpton by a household. Fast food s avalable n the market at a wde varety of prce and qualty ponts, but prce and qualty do not seem to nfluence consumpton frequency. Conclusons Gven changes n food consumpton behavors, there s a consderable need to dentfy the determnants of fast food consumpton. Ths research analyzed factors affectng the frequency of fast food consumpton among households n Mashhad, Iran. The top reasons for consumng fast food by households n a month n Mashhad were entertanment and lack of tme to prepare food at home. The results of estmatng negatve bnomal regresson (NBR) model show educaton, ncome, spouse occupaton, other meat consumpton, awareness about composton of fast food, avalablty of fast food restaurants, and advertsng varables have sgnfcant effect on the consumpton frequency of fast food by households. Consumer awareness about the composton of fast food has a negatve effect on the frequency of consumpton. Therefore, t s recommended that fast food supplers be proactve and provde healther fast food as a strategc response to consumers concerns n order to ncrease sales and profts. Qualty assurance provsons through socal meda could ncrease populaton awareness and demand for healthy foods. Polcy makers could use socal marketng tools to create ncentves for households to eat more healthy products. It s also recommended that tradtonal Iranan foods, whch have hgh nutrtonal values and are quck to prepare, to be added to the menus of fast food restaurants. Acknowledgments Ths s publcaton No. 17-04-105 of the Kentucky Agrcultural Experment Staton and s publshed wth the approval of the Drector. The Natonal Insttute of Food and Agrculture, U.S. Department of Agrculture, Hatch project under KY004043 0229745 supported ths work. March 2018 27 Volume 49, Issue 1
References Afolab, W., O. Oyawoye, S. A. Sann, and O. O. Onabanjo. 2013. Proxmate and Cholesterol Composton of Selected Fast Foods Sold n Ngera. Ngeran Food Journal 31(1):70 76. Austn, S. B., S. J. Melly, B. N. Sanchez, and A. Patel. 2005. Clusterng of Fast Food Restaurants around Schools: A Novel Applcaton of Spatal Statstcs to the Study of Food Envronments. Amercan Journal of Publc Health 95(9):1575 1582. Carns, J. E., and S. R. Rundle-Thele. 2014. Eatng for the Better: A Socal Marketng Revew (2000 2012). Publc Health and Nutrton 17(07):1628 1639. Chrstan, M., and G. Gereff. 2010. The Marketng and Dstrbuton of Fast Food. In M. S. Freemark, ed. Pedatrc Obesty. New York, NY: Sprnger, pp. 439 450. Fannng, J., T. Marsh, and K. Stegert. 2002. Determnants of Fast Food Consumpton, Paper presented at WAEA annual meetng, Long Beach, Calforna, July 28 31. Hearst, M. O., L. G. Harnack, K. W. Bauer, A. A. Earnest, S. A. French, and J. M. Oakes. 2013. Nutrtonal Qualty at Eght U.S. Fast-Food Chans: 14-Year Trends. Amercan Journal of Preventve Medcne 44(6):589 594. Lee, S., and N. Len. 2015. The Influence of Adult Famly Members on Chldren's Fast Food Consumpton: A Health Belef Perspectve. Journal of Communcaton n Healthcare 8(3):185 196. Lee, N., and P. Kotler. 2015. Socal Marketng: Changng Behavors for Good, 5th ed. Thousand Oaks, CA: Sage. Long, S. J. 1997. Regresson Models for Categorcal and Lmted Dependent Varables. London: Sage. Long, S. J., and J. Freese. 2001. Regresson Models for Categorcal Dependent Varables Usng Stata. College Staton, TX: Stata Press. Omar, R., and G. Frempong. 2016. Food Safety Concerns of Fast Food Consumers n Urban Ghana. Appette 98:49 54. Park, C. 2004. Effcent or Enjoyable? Consumer Values of Eatng-Out and Fast Restaurant Consumpton n Korea. Internatonal Journal of Hosptalty Management 23:87 94. Sahagun, M. A., and A. Z. Vasquez-Parraga. 2014. Can Fast Food Consumers Be Loyal Customers, If So How? Theory, Method and Fndngs. Journal of Retalng and Consumer Servces 21:168 174. March 2018 28 Volume 49, Issue 1
Song, Y. 2016. Factors That Affect Fast Food Consumpton: A Revew of the Lterature. Provdence, RI: Johnson& Wales Unversty, MBA Student Scholarshp, Paper 53. Avalable onlne: http://scholarsarchve.jwu.edu/mba_student/53 World Health Organzaton. 2015. Q&A on the Carcnogencty of the Consumpton of Red Meat and Processed Meat. Avalable onlne: http://www.who.nt/features/qa/cancer-redmeat/en/ Xue, H., Y. Wu, X. Wang, and Y. Wang. 2016. Tme Trends n Fast Food Consumpton and Its Assocaton wth Obesty among Chldren n Chna. PLoS One 11(3): e0151141. March 2018 29 Volume 49, Issue 1