Status Inconsistency and Drinking Patterns among Working Men and Women Douglas A. Parker. Ph.D.. Elizabeth S. Parker, Ph.D.. Thornas'C. Harford, Ph.D., and Jacob A. Brody, M.D. Previous research suggests that status inconsistents constitute a social category that is vulnerable to embarrassments and disappointments, since individuals who are of high status on one dimension and low status on another tend to respond to themselves in terms of the higher status, while other people tend to respond to them in terms of the lower one. This article extends this research to patterns of alcohol consumption and finds that status and status inconsistency are significantly associated with frequency of consumption and quantity consumed per occasion among certain groups of drinkers.

F

OR MORE THAN a quarter of a century, sociologists and others have been concerned with the products of status inconsistency. Preferences for political parties, liberal attitudes on economic issues, symptoms of physical and mental illness, and forms of coronary heart disease have been associated with status inconsistency.1-5 This article extends inconsistency research to drinking behavior. Drawing on recent conceptual and technical innovations,6 we examined the relationships between status inconsistency and drinking patterns among employed men and women in metropolitan Boston. Status inconsistency has been defined as the occupancy of positions at different levels of ranked status orders.' An often cited example of a status inconsistent is the black physician who is thought to have a low ethnic status and a high occupational status.' It is held that status inconsistents constitute a social category that is vulnerable to embarrassments and disappointments, since individuals who are of high status on one dimension and low status on another tend to respond to themselves in terms of the higher status, while other people tend to respond to them in terms of the lower Previous research suggests that there are two forms of inconsistency. The first type, which we will call ascription inconsistency, is manifested when problematic behavior increases as ascribed status, such as ethnic position, exceeds

achieved status, such as occupational level. The second type, which we will call achievement inconsislency, is evidenced when problematic behavior increases as achieved status exceeds ascribed status. Research also suggests the psychologic components that mediate the inconsistency effects: ascription inconsistents will have a sense of personal failure and feelings of self-blame, and achievement inconsistents will have a sense of deprivation and an inclination to blame other^.^.^ These properties of status inconsistency are summarized in Table 1. This article presents an examination of the relationships between alcohol 'consumption and the stressful conditions of status and status inconsistency. Drinking has long been viewed as a response to stresses and strains in the environmen t, presumably because of its t ension-reducing properties. l o Indeed, several types of environmental s t r e s s have been found to be associated with increased drinking, and these include subsistence anxiety," economic anxieties,12 and unemployment and reduced income. n Recent laboratory investigations have shown that the alcohol consumption o,f heavy social drinkers increased significantly in the face of the threat of evaluation by others.'* A less ego-involving stress comprised of threat of an electric shock had no significant effect on the alcohol consumption of alcoholic and nonalcoholic drinkers,I5 suggesting that situations involving the stress of interpersonal evaluation are related to increased alcohol consumption." Since status inconsistency is stressful, in part because the inconsistents' evaluation of self is at odds with the evaluation of others, it may well be a significant predictor of increased drinking behavior. From the Clinical and Biobehavioral Branch and the Epidemiological and Special Studies Branch, National Instituteon Alcohol Abuse and Alcoholism. Rockville. Md. Reprint requests should be addressed to Douglas A . Parker, Ph.D.,Epidemiological and Special Studies Branch. National lnsriture on Alcohol Abuse and Alcoholism. 5600 Fishers Lane. Rockville. Md. 20857. @ 1978 by Grune & Stratton. Inc. 0145-4008 178 J0202-OO27%0l.OOJO

Alcoholism: Clinical and Experimental Research, Vol. 2 . No. 2 (April). 1978

101

PARKER E l AL.

102

Table 1. Properties of Status Inconsistency

Example Definition

Psychologic component Statistical representation

Tvpe 1 Ascription Inconsistency

Achievement Inconsistency

White Laborers Problematic behavior increases as ascribed status exceeds achieved status. e.g.. drinking increases as ethnic position exceeds occupational level A sense of personal failure and feelings of self-blame Negative slope of inconsistency coefficients

Black Professionals Problematic behavior increases as achieved status exceeds ascribed status, e.g.. drinking increases as occupational level exceeds ethnic position A sense of deprivationand feelings of blaming others Positive slope of inconsistency coefficients

MATERIALS AND METHODS This report draws on data from a 1974 survey of metropolitan Boston.Ie A two-stage stratified probability sample of 795 men and women who were aged 18 and over and consumed alcohol at least once a month, were interviewed over a 4-wk time period. The analysis presented here is based on only the initial or “baseline” interview data for employed persons.

Status and Drinking Measures The primary status dimensions employed in the 1974 survey were ethnicity, occupation, education, and personal income. Only the first two status dimensions, ethnicity and occupation, are utilized in this report. For ethnicity, four rank positions were established by use of the Bogardus Social Distance Scale for the northeast.” Rank I, scale range 100-119, includes North Americans; rank 2, scale range 120-179. includes Northwestern Europeans; rank 3, scale range 180-239, includes Southern and Eastern Europeans and Japanese Americans; and rank 4. scale range 240-299, includes Afro-Americans, American Indians, Mexican-Americans, and Asians. For occupation, four rank positions were established by use of the Duncan Socioeconomic Index for Occupations.I8 Rank I, scale range 70-99, includes professionals, managers, and administrators; rank 2. scale range 50-69, includes technicians and clerical workers; rank 3, scale range 30-49, includes craftsmen and foremen; and rank 4, scale range 1-29, includes operatives, service workers, and laborers. Drinking patterns were analyzed in terms of frequency and quantity of the use of beverage alcohol. Since there may be beverage-specific responses to fluctuations in the national economy,ls this article analyzes the drinking patterns of employed persons by wine, beer, and liquor to determine to what extent the status and status inconsistency effects are beverage-specific. Frequency of drinking was determined separately for wine, beer, and liquor, and scores could range from every day to never. Similarly, for each occasion in which a particular beverage was consumed, the average quantity of beverage consumed was determined, and scores could range from 12 or more glasses of wine, cans of beer, or drinks of liquor to less than 1.

A n Additive Measurement Model Establishing the strength and direction of inconsistency effects has involved serious methodological difficulties. The fundamental problem is that of distinguishing between the

Type 2

effects of inconsistencies between two or more social positions and the effects of the positions themselves, that is, between inconsistency effects and status effects. Previous research attempted to test for an inconsistency effect by first computing the regression of a dependent variable on two status variables and then adding a discrepancy term to the equation. The difficulty is that estimates yielded by the second equation are identical to those yielded by the first. The covariance matrix for the three terms is singular, and thus, the regression coefficients cannot be uniquely determined. In a recent paper,” Hope argues that the step which invalidates the model of previous research is the assumption that the parameters of the status dimensions are to be separately estimated from the data. Hope points out that early formulations of status inconsistency theory define inconsistency not by contrast to the separate effects of the status variables but rather by contrast to the shared effects. The theory specifies that some aspects of each of the status dimensions should be controlled but not all aspects of all dimensions. Hope suggests that status inconsistency is present to the degree that the estimates obtained from the regression of a dependent variable on the set of independent variables (each contributing its separate and distinct effect) fit the data better than estimates obtained by the regression of the dependent variable on the dimension of shared effects. Hope’s model is employed here in the analysis of drinking patterns.

Equations and Statistical Method A fairly straightforward application of stepwise multiple regression procedures was made to test for the effects of status, inconsistency, and the interaction between status and inconsistency. We estimated three standardized regression equations in a predetermined 0rder.l’ Equation I is for status effects only, equation 2 is for status and inconsistency effects, and equation 3 is for status, inconsistency, and interaction effects as follows: DP = BIXl + e DP

=

BIXl + BiXa

DP = BIXl + BzXa

(1)

+e +

BaX3

(2)

+e

where DP is drinking pattern score, and XI = Z ,

(3)

+ Z2 - 1,

X2= ZI - Z2, and X 3 = XIXa, where ZI is ethnic rank

score, Z, is occupational rank score, and e is residual variance.

STATUS INCONSISTENCY AND DRINKING PATrERNS

103

The predetermined order of variables in the standardized regression equations is dictated by sociologic research.'l Because status inconsistency is not possible without status, the status term is always entered first and retained in equations 2 and 3 regardless of whether or not it significantly accounts for a proportion of the variance. This constitutes a conservative test of inconsistency effects because we are allowing the status term to claim credit for the direct effect of status on drinking behavior as well as the indirect effect through status inconsistency (that is, for their commonality). To be persuaded that it is worth adding an inconsistency term to the prediction equation for drinking patterns, we require the following: (A) that the standardized regression coefficient for inconsistency be significant, (B) that the squared multiple correlation coefficient be significant, and (C) that the increment to the variance accounted for exceed 1% of the total variance.z3The final requirement is imposed because if equation I, the simpler equation containing only a .

Table 2.

status term, predicts equally well, then in the interest of parsimony we will discard the more elaborate equations. Any coefficient with a p >0.05 was not accepted as significant. None of the interaction coefficients was significant by this criterion, and therefore, equation 3 was dropped from the analysis.

RESULTS AND DISCUSSION

Status predicts drinking behavior, as the standardized regression coefficients in the second rows of fable 2 indicate. For certain groups of younger men and women, status is to the frequency Of consumption of wine, beer, and liquor as shown bY the P s i tive sign of the significant status coefficients.

Status and Status Inconsistency Effects on Drinking Patterns of Employed Personsby Beverage, Age, and Sex Women

Man

-

Aged L 35 IN * 96)

Aged 1a-34 IN 143)

Aged 2 35 IN * 130)

Aged 18-34 IN * 109)

.0458t .2063t ,1008 ,0559'

,0053 ,0775 .2574t .0715t

,0454' .221 6' ,1015 ,0563'

,0060 ,0806 - .0588 ,0095

.ooo3 ,0159 ,0358 ,0016

,0003 ,0138 - ,0648 .0045

,0392 ,1964 ,0648 .0434

,0016 -.0286 -.1458 ,0227

,0134 -.1185 -.1568 ,0380

,0375' ,1833' -.1249 ,0530'

,0283 -.1681 -.0184 ,0287

.oOOo

.0128 ,1099 - ,2259' .0638'

.moo -.0017 - ,0807 ,0065

,0011 ,0404 - .3670t ,1357'

.0571t .2232t .2056t .0992$

,0037 ,0625 .lo29 ,0143

,0277 ,1680 ,0181 ,0281

,0298 ,1734 - ,0782 .0356

.0560t -.2272t - ,0908 ,0641'

.0418'

- ,2060' - ,0709

,0070 ,0925 ,1092 ,0189

Wine frequency 1 2 3 4 Wine quantity 1 2

3 4

,0104 ,0992 .3213t ,1137t

Beer frequency

1 2 3 4 Beer quantity 1

2 3 4 Liquor frequency 1 2 3 4 Liquor quantity 1 2 3 4

- ,0088 ,1638 ,0266

.0469

1 : Proportion of variance accounted for by status effect (equation 1). 2: Standardizedregression coefficient for status effect (equation2). 3: Standardizedregression coefficient for inconsistencyeffect (equation 2). 4: Proportion of variance accounted for by status and inconsistencyeffects (equation 2). ' p < 0.05(F test). t p < 0.01(Ftest). $ p < 0.001 (Ftest).

,0044

- ,0624 -.1548 ,0283

104

For the younger groups of men and women, increasing status is significantly associated with increasing frequency of wine consumption, possibly reflecting the current vogue in wine drinking among younger age groups. For younger women, frequency of beer consumption is directly related to status, and this may also reflect a change of taste and the development of a new market in higher status groups. These results for younger women confirm the 1969 national survey finding of more drinking among young, higher status women.” For younger men, frequency of liquor consumption is directly associated with status. In contrast to this positive relationship between status and frequency, status is inversely related to quantity of liquor consumed in younger and older men, as shown by the negative sign of the status coefficients. In other words, although for younger men frequency of liquor drinking increases with higher status, for all men in this study the quantity of liquor consumed per occasion increased as status decreased. These findings indirectly support and specify previous research, indicating that wine and beer consumption increase during prosperity, and liquor consumption increases both with long-term prosperity and short-term economic stress. l 3 Status inconsistency also predicts drinking behavior, as t h e standardized regression coefficients in the third row of Table 2 indicate. The criteria for adding an inconsistency term to the prediction equation appear to be met. In a number of cases, significant relationships were observed and in all such cases the increment to the variance accounted for exceeded 1 % of the total variance as evidenced by the differences between the squared correlation coefficients in rows 1 and 4 of Table 2. For certain groups of men and women, discrepancies between status positions predict heavier and lighter drinking, and the types of discrepancies predict the particular drinking

PARKER ET AL.

pattern. The first type, ascription inconsistency, may produce feelings of personal failure, such as those that may be experienced by white laborers. These inconsistents may respond to their lack of achievement by some form of escape or binge drinking. The coefficients on Table 2 suggest that this pattern may be evident among older groups of men and women. Among those groups, as ethnic status exceeds occupational status, quantity of beer consumed increases; and, as occupational status exceeds ethnic status, beer quantity decreases. The other type, achievement inconsistency, may produce feelings of frustration, such as those that may be experienced by black professionals. These inconsistents may respond to the lack of recognition of their achievement by drinking more frequently though not more heavily. Such a pattern may be evident among older groups of men and women, since frequency of wine drinking increases as occupational s t a t u s exceeds ethnic s t a t u s and decreases as ethnic status exceeds occupational status. It may also be manifested in frequency of liquor consumption among younger men. Both types of inconsistency may constitute a threat to the self-esteem of persons. A sense of inadequacy can be produced by each form: in the first, because the inconsistents have not achieved; in the second, because other people have not recognized their achievements. That the types of inconsistency predict drinking behavior is taken to support the notion that structural variables can be important determinants of drinking behavior. We are continuing to explore the effects of status and status inconsistency on a wider range of dimensions in order to obtain a more complete picture of the social bases of alcohol consumption. ACKNOWLEDGMENT We are indebted to Karen Anderson, Dr. Marsha S. Harman. Henry Malin, and Charles Pautler for a number of helpful comments at earlier stages of the research.

REFERENCES 1. Lenski GE: Status crystallization: A non-vertical di5. Shekelle RB, Ostfeld AM, Oglesby P: Social status mension of social status. Am Sociol Rev 19:405-413, 1954 and incidence of coronary heart disease. J Chron Dis 2. Lenski G E Social participation and status crystalliza22~381-394,1969 tion. Am Sociol Rev 19:458-464. 1956 6. Hope K Models of status inconsistency and social mo3. Jackson EF: Status consistency and symptoms of bility erects. Am Sociol Rev 40:322-343, 1975 stress. Am Sociol Rev 27:469-480, 1962 7. Hornung CA: Social status, status inconsistency and 4. Gove WR: The relationshipbetween sex roles, marital psychological stress. Am Sociol Rev 42523-638. 1977 status, and mental illness. SOCForces 5 1:34-44, 1972 8. Scgal D: Status inconsistency, cross pressures, and

STATUS INCONSISTENCY AND DRINKING PATTERNS

American political behavior. Am Sociol Rev 34:352-359, 1969 9. Olsen ME, Tully JC: Socioeconomic-ethnic status in-

consistency and preference for political change. A m Sociol Rev 37:560-574, 1972 10. Vogel-Sprott M: Alcoholism and learning, in Kissin B, Begleiter H ( 4 s ) : The Biology of Alcoholism, vol 2. New York, Plenum, 1972 1 1 . Horton D: The functions of alcohol in primitive societies. Q J Stud Alcohol 4:199-320, 1943 12. Pearlin LI, Radabaugh CW: Economic strains and the coping functions of alcohol. Am J Sociol 82:652-663, 1976 13. Brenner MH: Trends in alcohol consumption and associated illnesses. Am J Public Health 65:1279- 1292, 1975 14. Higgins RL, Marlatt GA: Fear of interpersonal

evaluation as a determinant of alcohol consumption in male social drinkers. J Abnorm Psychol 84:644651, 1975 IS. Higgins RL, Marlatt GA: Eflects of anxiety arousal on the consumption of alcohol by alcoholics and social drinkers. J Consult Clin Psychol 41:426-433, 1975 16. Gerstel EK, Mason RE, Piserchia PV, et al: A pilot study of the social contexts of drinking and correlates. Final report submitted by the Research Triangle Institute to the

105

National Institute on Alcohol Abuse and Alcoholism, Contract No. HSM-42-73-110 (NIA), 1975 17. Bogardus ES: Racial reactions by regions. SOC Sociol Res 43:286-290, 1959 18. Duncan OD: A socio-economic index for all occupations. in Reiss AJ ( 4 ) : Occupations and Social Status. Glencoe, 111, Free Press, 1961, pp 109-138 19. Blalock HM: The identification problem and theory building: The case of status inconsistency. Am Sociol Rev 31:52-61, 1966 20. Blalock HM: Status inconsistency, social mobility,

status integration and structural effects. Am Sociol Rev 32:790-80 I , 1967 21. Nie NH, Hull CH, Jenkins JG, et al: Statistical Package for the Social Sciences (ed 2). New York, McGraw-Hill, 1975 22. Taylor HF: Linear models of consistency: Some extensions of Blalock’s strategy. Am J Sociol 78:1192-1215. 1973 23. Horan PM, Gray BH: Status inconsistency, disability and coronary heart disease. J Health SOCBehav 15:300-310, 1974 24. Cahalan D, Cisin IH, Crossley H M : American

Drinking Practices. Piscataway, NJ, Rutgers Center of Alcohol Studies, 1969

Status inconsistency and drinking patterns among working men and women.

Status Inconsistency and Drinking Patterns among Working Men and Women Douglas A. Parker. Ph.D.. Elizabeth S. Parker, Ph.D.. Thornas'C. Harford, Ph.D...
363KB Sizes 0 Downloads 0 Views