This Article Abstract Full Text (PDF) Services Download to citation manager PubMed PubMed Citation

Older Men and Older Women in the Arms of Criminal Law

Offending Patterns and Sentencing Outcomes

^a The Pennsylvania State University, University Park

Darrell Steffensmeier, Sociology Department, The Pennsylvania State University, University Park, PA 16802 E-mail: d4s{at}psu.edu.

	Abstract

TOP Abstract Introduction Methods Results Discussion References

 
Objectives. This study examines whether older defendants receive more lenient sentences compared with their younger counterparts and whether the effects of aging on sentencing outcomes manifests itself similarly across male and female offenders.

Methods. Using statewide data from Pennsylvania for 1990–94, logit models were used to assess the effects of aging on the in/out or incarcerative decision, and, ordinary least squares models were used to assess the effects on the length-of-term decision.

Results. Results show that older offenders of both genders were sentenced less harshly—they are less likely to be imprisoned than their younger counterparts and, if imprisoned, elderly defendants receive shorter prison terms. However, the elderly advantage was diminished in the case of drug offending, and the within-gender elderly advantage was found to be greater for males than for women.

Discussion. While these age differences in sentence outcomes appear to thwart norms of judicial impartiality, they also might reflect legitimate sentencing concerns of judges (in areas such as crime propensity, blameworthiness, and even the extra costs needed to jail older defendants). Therefore, an overall pattern of less severe punishment of older defendants (and/or female defendants) may still be warranted.

	Introduction

TOP Abstract Introduction Methods Results Discussion References

 
DRIVEN largely by the rapidly changing age structure of the U.S. population, in which the number of senior citizens (e.g., aged 65 and older) has become the fastest growing age group (U.S. Bureau of Census 1996), empirical research on "advanced age" as an important basis of social differentiation in American society is flourishing in sociology and the social sciences more generally (Moen 1996; Riley 1994). The centrality of age roles, including advanced age, to people's lives has been documented in numerous life-course studies of the economy, the family, the educational system, and the polity (Riley, Foner, and Waring 1988). So, too, in the field of law/criminology, there is a growing body of research on the effects of age on criminal offending and victimization (see Flynn 1996; McCarthy and Langworthy 1992; Steffensmeier 1987; Steffensmeier and Allan 1995). The research shows that older persons have much lower levels of criminal offending and somewhat lower levels of criminal victimization, relative to other age groups (other than the very young).

In contrast, there are only a few studies that examine the effects of age on the sentencing of older criminal defendants. (There is some ambiguity in the "aging and life course" literature, but researchers typically use either ages 60 and older or 65 and older to suitably distinguish the middle aged from those who are older or elderly [see review in Achenbaum 1985; see also Hagestad and Neugarten 1985; Pampel 1998]). These studies tend to support conventional wisdom, that is, the older defendant receives more lenient treatment in the arms of the criminal law. However, the studies in this area are marked by varied shortcomings such as the failure to simultaneously adjust for the effects of legal variables like prior record and offense seriousness (Champion 1987; Wilbanks 1988), which other research shows also influence sentencing outcomes. Additionally, the research to date has targeted older male while virtually ignoring older female defendants.

As a result, researchers know very little about whether older defendants are sentenced less harshly than their younger counterparts and whether the presumed lenient treatment of older defendants manifests itself similarly across male and female defendants. Conversely, we know very little about whether the greater leniency presumably accorded to female defendants exists across all age categories, including older defendants. The present study addresses these gaps in sentencing research by examining the independent and joint effects of age and gender in sentencing and by comparing the sentencing outcomes of older male and older female defendants to their younger counterparts.

Prior Research
Prior research on the independent effects of age on sentencing outcomes has been sparse. Most analyses of sentencing have included age as a control variable and treat it as a continuous variable, thus assuming a linear effect. These studies reported age as having a small, negligible effect on sentencing outcomes (Myers and Talarico 1987; Peterson and Hagan 1984). In contrast, several studies have found that when the data are subdivided into "old" versus "young" subgroups (e.g., aged 50 and older versus under age 50), denoting that older offenders tend to receive more lenient treatment than their younger counterparts. However, it is hazardous to draw firm conclusions from these studies because of shortcomings such as small sample size and the lack of statistical controls for prior record and offense seriousness (Champion 1987; Wilbanks and Kim 1984).

Importantly, some recent research suggests that the age–sentencing relationship is more complex than previously thought. When the full range of ages are partitioned into groups, an age effect is reflected in the form of an inverted U-shaped relationship from the late teens to young adulthood through mid-life and old age (Steffensmeier, Kramer, and Ulmer 1995). Net of controls for sentence severity and prior record, the youngest or late-adolescent age group (18–20) receives more lenient treatment at sentencing than does the young adult age group (21–29) and about the same leniency as those aged 30–39. Lenient treatment by judges increases for those from 40 to 49 years old. Defendants aged 50 years and older receive the greatest leniency.

In contrast to the scarcity of research on the age–sentencing relationship, there now exists a considerable amount of research on the effects of gender in the sentencing of adult defendants. The studies generally show that female defendants are treated more leniently than male defendants and that the gender differences are more apparent in sentencing or imprisonment decisions than in case dismissals or convictions (see reviews in Daly and Bordt 1995; and Steffensmeier, Kramer, and Streifel 1993). Largely because of the small number of female defendants, however, prior research has not examined the effects of age and especially the effects of older or senior citizen age defendants on the gender–sentencing relationship.

These age and gender findings suggest the importance of comparing sentencing outcomes across age–gender groupings and comparing, in particular, the sentencing outcomes of older male and older female defendants. Given the marked increase of elders relative to other age segments in the population, the lack of research on the older offender in the arms of criminal law is unfortunate. Even if their crime rates are comparatively small, the number of older offenders being sentenced in our criminal courts will continue to increase and their handling by those courts will assume greater theoretical and policy relevance.

Using statewide data from Pennsylvania for 1990–1994, we examined whether the older offender receives especially lenient treatment and whether the more lenient treatment of older defendants manifests itself similarly across male and female comparisons. The Pennsylvania data are exceptionally suited for this assessment because they include: (a) a large number of cases which cover a range of offenses, (b) rigorous controls of legal variables such as prior record and offense seriousness, and (c) the ability to distinguish sentencing outcomes in relation to two critical stages of sentencing—the decision of whether to imprison and length of term.

Theoretical Expectations
Prior-sentencing research suggests that judges are guided by three focal concerns in reaching sentencing decisions (Steffensmeier, Ulmer, and Kramer 1998) and that they typically have limited time and information about defendants when rendering their sentencing decisions. The three focal concerns are (a) the offender's blameworthiness and the degree of harm caused the victim, (b) the protection of the community, and (c) the practical implications of sentencing decisions. Blameworthiness is associated with offender's culpability and having the punishment fit the crime. Judges' views of blameworthiness are influenced mainly by offense severity, by offender's biographical factors such as criminal history (which increases perceptions of blameworthiness and risk) or prior victimization at the hands of others (which tends to mitigate perceived blameworthiness), and by the offender's role in the offense (e.g., being a leader or organizer increases blameworthiness). Protection of the community draws on similar attributions but focuses more on the need to incapacitate the offender or to deter would-be offenders. Judges' assessments about offenders' future behavior (dangerousness, recidivism) are based on attributions predicated on the nature of the offense (e.g, violent, property, drug), case information, the offender's criminal history, and also, perhaps, on characteristics of the offender such as education, employment, or community ties. Practical considerations come into play in sentencing decisions. These include concerns about the offender's "ability to do time," the costs to be borne by the correctional system, and the disruption of ties to children or other family members. Also, judges are likely to be concerned about the impact of offender recidivism on the court's standing in the public's eye and on their judicial careers.

The focal concerns and their interplay are complex, and judges rarely have complete information about cases or defendants. To reduce uncertainty, judges may rely not only on the defendant's present offense and prior criminal conduct, but also on attributions linked to the defendant's age and gender (Steffensmeier, Kramer, and Streifel 1993; Ulmer 1997). On the basis of these attributions or pieces of information, judges may project behavioral expectations about whether the offender (a) is likely to be a good or bad risk for rehabilitation; (b) is a potential danger to the community; (c) is more or less blameworthy, and thus more or less deserving of punishment; and (d) poses special practical considerations, such as inability to do time or organizational and staffing concerns about dealing with inmate health needs.

Hence, we expect that age and gender attributions will intertwine with the focal concerns outlined above to influence judges in deciding whether to incarcerate an offender and in determining the length of incarceration. Our guiding hypothesis is that older defendants will be sentenced more leniently than their younger counterparts but that this age effect will be conditioned somewhat by the defendant's gender and by the type of conviction offense. Specifically, four hypotheses frame our analysis.

Hypothesis 1: Older male defendants as well as older female defendants will be sentenced more leniently than their same-gender counterparts
The writings on aging generally show that, relative to individuals in other stages of the life course, individuals in childhood and old age are more likely to be evaluated on the basis of age than on performance criteria. One consequence of this is that invidious distinctions or expectations associated with criminal behavior are less likely to be attributed to older offenders—both in perception and in fact.

This hypothesis is based primarily on the expectation about differences between younger and older offenders in danger or threat to society (which contributes to harsher sentencing of younger defendants). Aging tends to soften attributions such as danger, unconventionality, and commitment to "street life" that contribute to harsher sentencing of younger defendants. The decline in physical prowess and pugnacity with advanced age leads to the expectation that older persons are less aggressive and less able to use force in threatening or harming others. Judges (and other court officials) may also perceive older offenders as less likely to recidivate—that is, better able to reform themselves and less likely to associate with members of a crime-generating peer group. The behavior of older offenders is more likely to be seen as idiosyncratic—that is, explained away as the result of forces outside of their control (e.g., extreme environmental circumstances, problems with health associated with advanced age). Finally, older offenders, from experience, may be more able to convince the judge of a more lenient sentence by showing remorse or reform (Steffensmeier et al. 1995).

A prison or jail sentence also may be perceived as presenting greater physical and psychological demands on the older inmate—older offenders may be seen as more vulnerable to aggression from younger offenders and in other ways may adapt less well to conditions of confinement. Time for older offenders is more likely to be seen as a diminishing, exhaustible resource in which the future becomes increasingly valuable. A year of imprisonment given to an offender in his or her 60s takes a considerably larger proportion of that person's remaining years than the same punishment imposed on a 20-year-old (Sherwin 1990). Moreover, judges may consider the practical matter of the costs associated with sentencing a senior citizen to prison—that is, as financially costly, burdensome to correctional officials, and posing special problems for prison staff such as special diets, medications, and assorted health problems (Goetting 1983).

Hypothesis 2: Gender effects will exist across all ages—that is, within each age group, including among older offenders, female defendants will be sentenced more leniently than male defendants
Women are expected to be treated more leniently than men, independent of the effects of age. Attributions of threat to society apply more to male than female offenders, especially younger male offenders. Steffensmeier et al. 1993 found that judges were less likely to jail female defendants, although "on paper" they had committed similar crimes. Judges reported accounting for such variation as due to gender differences in risk or danger to the community, in degree of blameworthiness, and in the practical effects of imprisoning women for both the correctional system and the community at large (Steffensmeier et al. 1993). Judges were concerned about social costs to children of sending women to prison and the organizational demands of imprisoning pregnant women or women with other physical or mental health problems.

Hypothesis 3: Age effects and the older-aged advantage will be greater among male defendants than among female defendants (i.e., within-gender differences in age effects will be greater among male defendants)
Differences between younger and older offenders in danger or threat to society are expected to be held more strongly in reference to male than female offenders. Issues such as risk to community or propensity for crime will be more affected by aging for male than female offenders. Also, differences in indicators of conventionality and responsibility for others are greater between younger and older men than between younger and older women. In fact, many younger female defendants might benefit from having children as dependents when it comes to sentencing.

Hypothesis 4: Age effects and the older-aged advantage in sentencing outcomes will be contextualized by type of offense, notably drug offending
Some prior research and anecdotal evidence suggests that the age effect is diminished for drug offenders—apparently because the drug behavior is viewed as a high risk factor for recidivism and perhaps because older drug offenders are viewed as less amenable to treatment and because society has no choice but to lock them up (Pennsylvania Crime Commission 1991; Steffensmeier et al. 1995). Also, drug dealing by older defendants is likely to be seen as especially repugnant and irresponsible, in particular, because their clientele is likely to be younger and presumably less mature (Steffensmeier et al. 1995). Therefore, the elderly advantage in sentencing outcomes is expected to be less for defendants convicted of drug offenses than for those convicted of violent and/or property offenses.

	Methods

TOP Abstract Introduction Methods Results Discussion References

 
The data used in this article are from the Pennsylvania Commission on Sentencing (PCS) database for the years 1990–1994. By law, each sentence given for a felony or misdemeanor conviction must be reported to the PCS. The data provide detailed information on sentences given, as well as unusually complete information on offense severity, prior record, and other offender-related and court contextual variables that might affect sentencing. The data also provide an unusually large number of cases (close to 175,000, including over 8,000 defendants aged 50 and older). Because we target the effects of advancing age on sentencing and also to smooth out the curvilinear age effect described earlier (i.e., more lenient sentencing of 18–20-year-olds compared with young adults in their 20s and 30s), we limit our analysis to adult defendants aged 21 and older.

Pennsylvania implemented presumptive guidelines in 1982 to structure, but not eliminate, sentencing discretion. Guideline sentence ranges are established for each combination of offense severity and criminal history in the form of a sentence matrix (see Ulmer 1997). However, because there is considerable spread in the ranges, judges can still exercise a fair amount of sentencing discretion. Our measure of prior record is a 7-point scale of prior convictions, weighted according to their severity, which includes prior misdemeanors punishable by at least 1 year's incarceration and all prior felonies. Prior misdemeanors may total no more than 2 points on the scale, whereas prior felonies count for 1 to 3 points each, depending on their severity. Our measure of offense severity involves a 9-dummy-coded gravity score for nondrug offenses and a 9-dummy-coded score for drug offenses (i.e., one for each level of severity) that are based on scales developed by the PCS that rank each statutory offense. The severity scales subdivide offenses in terms of severity on the basis of such factors as degree of victim harm, offender culpability, weapon use, property loss, or drug amount. Regarding nondrug offenses, for example, a severity score of 9 is assigned to offenses such as forcible rape or robbery committed with a firearm, whereas a severity score of 3 is assigned to offenses such as criminal trespassing or check forgery amounting to less than $250. Regarding drug offenses, for example, a severity score of 9 is assigned to offenses such as the sale or distribution of 100+ grams of narcotics (i.e., cocaine or heroin), whereas a score of 3 is assigned to the sale or distribution of small amounts marijuana (i.e., less than one pound).

Table 1 shows the independent and dependent variables that we used in this analysis and describes their coding. Age of the defendant is the central independent variable and is defined as the age of the offender at sentencing. Age groupings are created as follows: ages 21–29, ages 30–39, ages 40–49—and then to show more pointedly the effects of aging—ages 50–59, and ages 60 and older. Moreover, in cases in which sample size was adequate, we extended the age groupings to include ages 50–54, 55–59, 60–64, 65–69, 70–74, and 75 and older. Age is also used in conjunction with gender resulting in five age–gender combinations (21–29, 30–39, 40–49, 50–59, and 60+); these comparisons are capped at 60+ to ensure adequate sample size, especially in view of the small number of older female offenders. We also included other court contextual variables identified by other studies as influencing sentence outcomes such as mode of conviction, county percentage Republican voters, county uniform Crime Report Index rate, and defendant's race (Steffensmeier et al. 1995; Ulmer 1997). A time variable is also included to control for variation by year.

Although the Pennsylvania guideline data provide some of the richest information in the country for analyzing sentencing outcomes, several caveats are in order. First, we examine only the sentences imposed on convicted offenders. Thus, we do not address whether Age x Gender differences exist in earlier stages of processing or prosecution (e.g., pretrial release, charging decisions). Nonetheless, because it is virtually impossible to account for all of the selection processes (both formal and informal) that operate in the criminal justice system, some degree of uncorrected sampling bias is common in every study. Second, some degree of measurement error is likely to be present in our indicators for legal and extralegal variables and may be greater in the legal variables, thus attenuating their effects relative to demographic variables such as age and gender. Third, although we control for important contextual variables, the inclusion of these variables in our analysis is far from exhaustive.

	Results

TOP Abstract Introduction Methods Results Discussion References

 
Table 2 shows the offense distribution for all defendants convicted in Pennsylvania courts from 1990–1994, including a breakdown by gender. Displayed in columns 4–9 are the distributions for older male and female defendants, subdivided into ages 50–59, 60–69, and 70+ (i.e., each age group's percentage of total convictions and percentage of total same-gender convictions). Columns 10 and 11 show the offending profiles of older men and women defendants (i.e., they represent the percentage of all older convictions within each gender that are convictions for that particular offense).

Table 2 (columns 10, 11) also shows the offending profiles of older male and older female defendants, that is, the percentage of all older-age convictions within each gender that are convictions for that particular offense. The homicide figures of 11.8 for men and 5.7 for women indicate that about 12% of all male convictions involving persons aged 60 and older were for homicide, as compared with only 6% of all female convictions. Examining these percentages further, we find (a) property convictions are the most frequent offense types for both older men (22%) and older women (57%); (b) about two-thirds of the property-crime convictions involving older women are for retail theft offenses (felony and misdemeanor); (c) along with property convictions, older women are most often convicted of other misdemeanor offenses (14%) followed by serious drug offenses (11%) and violent offenses (6%); (d) for males aged 60 and older, the offense types most common after property convictions include other misdemeanor offenses (18%), followed by miscellaneous offenses (16%), and sex offenses (15%). Overall, these patterns are consistent with findings from analyses of arrest statistics that also show a profile of older offenders as mainly committing minor property and public order offenses (Steffensmeier 1987).

Sentencing Outcomes
We turn next to the main focus of our analysis—the sentencing of older defendants, including the interaction of age and gender, as influences in sentencing. Bivariate correlations of the variables included in the analysis revealed that offense severity and prior record have large effects on sentence outcomes and thus are important statistical controls for estimating age effects. We also found that both age and gender are related to sentence outcomes—older defendants and women are less likely to be incarcerated and receive shorter sentences. But the latter also have lower prior record and offense gravity scores. At issue here, therefore, is whether the age advantage or the gender advantage persist when these and other variables known to influence sentencing are taken into account.

As noted earlier, we used logistic regression to analyze the in/out or probation/prison decision and OLS regression to analyze the length-of-term decision. Because our data set is not a sample, but rather contains all reported sentences with complete data (i.e., all U.S. citizens convicted in Pennsylvania courts over the 1990–94 period), statistical tests of significance do not apply in the conventional sense of assessing error in making inferences to the universe from which the sample was drawn (Blalock 1981; Raferty 1995). Also, because the number of cases included in our analysis is so large, many small sentencing differences between groups or categories often turned out to be significant in a statistical sense. Therefore, we place more emphasis on the direction and magnitude of the coefficients than on statistical significance levels and judge the relative importance of the substantive effects of the independent variables according to probability differences in the likelihood of incarceration and differences in months for sentence length. To compute probability effects, we use the formula, [(odds ratio)/(odds ratio + 1)] - .5. Nonetheless, even with a data set as large as this one, some cell sizes can become very small—particularly among the smaller female group—and suggest the need to pay some attention to statistical significance levels. Table 3 shows the results from models of the in/out and length-of-term decisions for age and gender. Briefly, we find that prior record and offense seriousness continue to have strong effects on sentence outcomes in our multivariate models. Those convicted of more serious offenses and those with more extensive and severe prior convictions are more likely both to be incarcerated and to receive a longer prison sentence. For example, when the odds (about 1.5) for criminal history are converted to probability differences, each additional increase in the criminal history score raises the probability of incarceration (on average) by 10% across each age group; sentence length is increased by about 6 months for each one-unit increase in criminal history score. (The effects of the dummy-coded offense categories are available on request from the authors.) Also, type of conviction (bench and jury trials vs. guilty pleas) is quite influential in the two sentencing outcomes. Apparently, defendants who opt for a trial—at which they are then found guilty—are at risk of judges imposing a "trial penalty," because trials are time consuming and expensive (Eisenstein and Jacob 1977; Eisenstein, Fleming, and Nardulli 1988; Ulmer 1997). Also, defendants who are convicted by way of a trial may be seen as lacking in remorse and as unwilling to accept responsibility for their law violation (Ulmer 1997). Regarding the other control variables, their effects on sentence outcomes are small or negligible and have little bearing on the Age x Gender findings presented here. The reference category for assessing age effects is young adult offenders, ages 21–29, who generally receive harsher sentences than other age groups. The results for the in/out decision are displayed on the left and those for the length of term decision are on the right side of Table 3 .

Fig. 1 also displays the age–sentencing relationship and documents further the key finding that age effects are most evident for defendants in their 60s and 70s. On average, the probability of defendants in their 60s being incarcerated is about 25% less than defendants in the 21–29-year-old group; and, if incarcerated, the older defendants received incarceration sentences on average 8 months shorter. The differences are even larger for defendants in their 70s—they were about 30% less likely to be incarcerated than 21–29-year-olds and received prison terms about 13 months shorter.

View larger version (12K): [in this window] [in a new window]   Figure 1. Age Differencesa in Probability of Imprisonment and Length of Term (in Months), Net of Controls. aAge group 21–29 is the reference category. bThe formula—[odds/(1 + odds)] - .5—is used to calculate probability differences across age groups in the likelihood of receiving an "in" or prison sentence. The odds ratios are shown in Table 3 and reflect the main effects of age, net of controls. c"Length of term difference" is the difference in length of prison term received (in months) compared with the reference group (age 21–29). These are the ordinary least squares (OLS) coefficients from Table 3 , reflecting the main effects of age, net of controls.

Importantly, the age effect or older-aged advantage is comparatively greater among male than female defendants. For total convictions as well as across the three offense groupings, and for both the in/out and term-length decisions, the within-gender difference between younger and older defendants is consistently greater among male defendants. This patterning especially holds among those aged older than 60, as older male defendants receive considerably more lenient sentences relative to their younger male counterparts than is the case when older female defendants are compared with younger women. For example, among men, the lesser probability of imprisonment for older defendants relative to young adult defendants is 33% (violent offenses), 31% property offenses, and 19% (drug offenses); among women, the percentage differences between old and young defendants, respectively, are 20%, 22%, and 13%. A similar pattern prevails for the term decision in which, among men, older defendants receive sentences 13 months and 14 months shorter than young adult defendants, respectively, for violent and property offenses; whereas, among women, the differences are 10 months for the violent offenses and 7 months for the property offenses. Taken together, these findings support our earlier position that attribution differences between younger and older offenders—as regarding dangerousness, recidivism risk, and commitment to street life—are more salient in comparisons between younger men and older men than between younger women and older women.

Age by Gender Rank Ordering on Sentence Outcomes
To further illuminate the interaction of age and gender in sentencing and also to summarize the older-aged advantage and its conditioning by gender, we estimated models of incarceration and length that included all the previous control variables plus dummy variables representing 10 age–gender subgroup combinations. Table 5 shows how gender and age combine to influence the odds of incarceration and sentence lengths, and demonstrates a rank ordering of age–gender categories from most to least severe sentence.

One final and very important point is demonstrated in Table 5 . Our analysis shows that differences in sentencing outcomes are especially striking when one compares the extremely different age–gender categories; for example, older women are about 30% less likely to be incarcerated and, when incarcerated, receive prison terms about 16 months shorter than those of young adult men. In sum, an analysis of the main effects of gender, and even of age, may conceal very substantial differences in sentencing outcomes when comparisons are ignored between the most dissimilar age–gender comparisons.

	Discussion

TOP Abstract Introduction Methods Results Discussion References

 
Our results generally support the age and the Age x Gender bias models articulated earlier. In addition to the strong effects of prior record and offense conduct, older offenders were less likely to be imprisoned than were younger offenders; and if imprisoned, older offenders received shorter prison terms. Notably, defendants in their 60s and 70s appeared to benefit the most from the overall greater leniency extended to older defendants. Also, as expected, this age advantage was greater in cases involving property and violent offending than in drug cases. Moreover, the above patterns were reflected among both male and female defendants, but the elderly advantage—when within-gender differences were examined—was greater among male defendants. Apparently, the effects of age-based expectations (e.g., lesser aggression and strength of older persons) in softening the "dangerousness" and "risk to the community" labels typically associated with criminal behavior apply more to male than to female defendants (see Steffensmeier et al. 1995).

The fact that the elderly advantage in sentencing outcomes is diminished for drug-type offenses suggests that judges are as likely (or nearly so) to attribute stability of disposition to commit future drug-law violations to older as compared with younger drug offenders. Judges also may be particularly disparaging of the corrupting leverage of older drug traffickers whose offense behavior departs too far from age-linked expectations to justify leniency. However, an important caveat here is that our findings of a diminished elderly advantage in drug cases must be interpreted cautiously in view of the small number of older persons convicted of drug offenses, which can contribute to highly unstable estimates. Nonetheless, it is consistent with our theoretical expectations as outlined earlier.

Taken together, our findings reinforce the focal concerns theory of sentencing as described earlier (see Steffensmeier et al. 1998) and also suggest that many similar interpretations underlie age and gender differences in sentencing. For example, younger offenders and male defendants, on the one hand, appear to be seen as more of a threat to the community, as less reformable, and as more capable of "doing time." Women and older offenders, on the other hand, are more likely to be seen as having ties to the community and to benefit from indicators of stability and conventionality such as employment or care of others. They also may be seen as potentially presenting greater costs and problems for the correctional system in terms of health care and child welfare. Additionally, the blameworthiness of women and older offenders is more likely to be mitigated by prospects of being victimized themselves—for example, by coercion at the hands of men (for women), drug and alcohol problems, or psychological disorders (see Steffensmeier et al. 1995; Wilbanks 1988).

Finally, our key finding—that older defendants (especially those in their 60s and 70s) who face criminal penalties receive less harsh sentences and are an advantaged group relative to defendants in younger age groups—documents the centrality of age criteria in our social system and reveals how formal distinctions among individuals are often based more on age than on behavior or performance (see Moen 1996; Riley 1994). This finding also raises an interesting policy question. Are the statistically-observed age disparities found among male as well as female defendant populations warranted or unwarranted? On the one hand, in view of the emphasis on equality before the law in our justice system, it is reasonable to expect that judges will sentence strictly on the basis of what defendants have allegedly done, not on who they are or on how old they are. Older offenders should receive sentences as severe (or as lenient) as younger offenders convicted of identical offenses and with similar prior records. On the other hand, it is arguable that age-based differences in crime propensity, blameworthiness, and even factors such as the "costs" to the justice/correctional system of jailing older offenders are legitimate considerations in judges' sentencing decisions. If so, then the overall pattern of less severe punishing of women and older offenders may still be viewed as warranted. Whatever is advocated, these two positions represent conflicting views of fairness and constitute value judgments not easily resolvable by empirical inquiry.

	Acknowledgments

 
We appreciate the helpful comments on earlier drafts of this paper by the reviewers and editor.

Received for publication April 22, 1999. Accepted for publication November 3, 1999.

	References

TOP Abstract Introduction Methods Results Discussion References

 

• Achenbaum W., 1985. Societal perceptions of the aging and the aged. Binstock R., Shanas E., , ed.Handbook of aging and the social sciences 2nd ed. 129-148. Van Nostrand Reinhold, New York.
• Berk R., 1983. An introduction to sample selection bias in sociological data. American Sociological Review 48:386-398.
• Blalock H., 1981. Social statistics McGraw-Hill, New York.
• Champion D., 1987. Elderly felons and sentencing severity: Intergenerational variations in leniency and sentencing trends. Criminal Justice Review 12:7-14. [Abstract]
• Daly K., Bordt R. L., 1995. Sex effects and sentencing: An analysis of the statistical literature. Justice Quarterly 12:141-175.
• Eisenstein J., Fleming R., Nardulli P., 1988. The contours of justice: Communities and their courts Little, Brown and Company, Boston.
• Eisenstein J., Jacob H., 1977. Felony justice Little, Brown and Company, Boston.
• Flynn E., 1996. Age and crime. Encyclopedia of gerontology, Vol. VI 353-359. Academic Press, San Diego, CA.
• Goetting A., 1983. The elderly in prison: Issues and perspectives. Journal of Research in Crime and Delinquency 20:291-309. [Abstract]
• Hagestad G., Neugarten B., 1985. Age and the life course. Binstock R., Shanas E., , ed.Handbook of aging and the social sciences 2nd ed. 35-61. Van Nostrand Reinhold, New York.
• McCarthy B., Langworthy R., 1992. Offender perspectives in criminology and criminal justice Praeger, New York.
• Moen P., 1996. Gender, age, and the life course. Binstock R. H., George L. K., , ed.Handbook of aging and the social sciences 3rd ed. 171-187. Academic Press, San Diego, CA.
• Myers M. A., Talarico S. M., 1987. The social contexts of criminal sentencing Springer-Verlag, New York.
• Pampel F., 1998. Aging, social inequality, and public policy Pine Forge Press, Thousand Oaks, CA.
• Pennsylvania Crime Commission1991. Organized crime in Pennsylvania: 1990 report Commonwealth of Pennsylvania, Conshohocken, PA.
• Peterson R. D., Hagan J., 1984. Changing conceptions of race: Towards an account of anomalous findings of sentencing research. American Sociological Review 49:56-70.
• Raferty A., 1995. Bayesian model selection in social research. Sociological Methodology 25:111-163.
• Riley M. W., 1994. Aging and society: Past, present and future. The Gerontologist 34:436-446. [Medline]
• Riley, M. W., Foner, A., & Waring, J. (1988). In N. Smelser (Ed.), Handbook of sociology (pp. 243–290). New York: Macmillan.
• Sherwin R., 1990. Employing life expectancy as a guideline in sentencing criminal offenders: Toward a humanistic proposal for change. Prison Journal 70:12-16.
• Steffensmeier D., 1987. The invention of the new senior citizen criminal. Research on Aging 9:281-311. [Abstract]
• Steffensmeier D., Allan E., 1995. Gender, age, and crime. Sheley J., , ed.Handbook of criminology 83-114. Wadsworth, Belmont, CA.
• Steffensmeier D., Kramer J., Streifel C., 1993. Gender and imprisonment decisions. Criminology 31:411-446.
• Steffensmeier D., Kramer J., Ulmer J., 1995. Age differences in sentencing. Justice Quarterly 12:583-602.
• Steffensmeier D., Ulmer J., Kramer J., 1998. The interaction of race, gender, and age in criminal sentencing: The punishment cost of being young, black and male. Criminology 36:763-797.
• Ulmer J., 1997. Social worlds of sentencing: Court communities under sentencing guidelines SUNY Press, Albany, NY.
• Wilbanks W., 1988. Are elderly felons treated more leniently by the criminal justice system?. International Journal of Aging 26:275-288.
• Wilbanks W., Kim P., 1984. Elderly criminals University Press of America, New York.
• U.S. Bureau of the Census1996. Current population reports U.S. Government Printing Office, Washington DC.

This Article Abstract Full Text (PDF) Services Download to citation manager PubMed PubMed Citation