音声ブラウザご使用の方向け: SKIP NAVI GOTO NAVI

THE ROLE OF COMPUTER SKILLS IN EMPLOYMENT AND EARNINGS FOLLOWING A SPINAL CORD INJURY

Douglas Kruse, Ph.D.
School of Management and Labor Relations
Rutgers University
New Brunswick, NJ, 08903
Telephone: 908-445-5991
Fax: 908-445-5991
Internet: dkruse@rci.rutgers.edu

Alan Krueger, Ph.D.
Department of Economics
Princeton University
Princeton, NJ, 08544
Internet: akrueger@pucc.princeton.edu

Susan Drastal, M.S.
Research Department
Kessler Institute for Rehabilitation
1199 Pleasant Valley Way
West Orange, NJ 07052

Web Posted on: December 8, 1997


INTRODUCTION

What are the employment and earnings effects of a severe disability? Do computer skills help cushion the economic losses often associated with disabilities? This study investigates the labor market effects of a severe traumatic disability, a spinal cord injury (SCI) in the neck or back, which typically causes paralysis below the level of the injury and requires the use of a wheelchair. About 250,000 people in the U.S. have SCI's, and there are about 10,000 new SCI's each year (primarily from vehicular accidents, sports and diving accidents, and falls). SCI individuals account for over one-third of wheelchair users under age 65.

We compare the employment experiences of a sample of people with SCI's to those of former co-workers before and after the injury, and to the current employment situation of the general population. The analysis is based on a 1994 telephone survey of New Jersey adults who have had SCI's within the past ten years. The survey sample consists of 391 SCI respondents and 71 paired former co-workers. In addition, we draw upon administrative earnings records for 463 people with SCI's (including both survey respondents and non-respondents) and 5073 other New Jerseyans over the 1985-94 period. We also make comparisons to the general population using Current Population Survey data from the U.S. Bureau of Labor Statistics.

The characteristics of the SCI sample closely match those in a nationally-representative sample of SCI individuals (Berkowitz et al, 1992). The sample is evenly split between quadriplegics (those with an injury in the neck, typically impairing the use of arms as well as of the lower body) and paraplegics (those with an injury in the back). Over two-thirds (69%) are men, reflecting a greater likelihood for men to engage in activities that can lead to SCI's (e.g., riskier sports, driving, and jobs). Almost one-third of the injuries (31%) resulted from vehicular crashes, while substantial numbers were due to sports or diving accidents (14%) and falls (15%). A minority received other long-term injuries with the SCI (12%), while over two- thirds (70%) are able to drive a car or van (which is possible with special adaptations to the car or van). The educational levels are similar to those in the general population, but the percent who are married is lower (consistent with other findings that SCI's decrease marriage rates and increase divorce probabilities)(Trieschmann, 1988; Stover et al., 1995). Close to three-fourths (73%) of the SCI sample receive some form of disability income (through Social Security Disability Income, Supplemental Security Income, Workers Compensation, or private insurance).


EMPLOYMENT, EARNINGS, AND JOB CHARACTERISTICS

What are the employment and earnings consequences of an SCI? While nearly all (93%) of the respondents in the sample had pre-injury employment, and 81% were employed at the time of the injury, only 40% have had any post-injury employment, and only 30% are currently employed (20% full-time). These estimates are consistent with previous studies of people with SCI's (Berkowitz et al., 1992). Most of those with post-injury employment had to switch occupations and/or employers. Of all who were employed before the injury, only one-sixth (16.9%) returned to the same job after the injury, and another one-ninth (11.2%) say they could have returned to the same job if additional training and adaptive devices had been available. The majority of these injuries occurred before the passage of the Americans with Disabilities Act (ADA); it remains to be seen whether the ADA will make a difference in employer accommodations so that people with SCI's can more easily return to their old jobs or find new ones.

Those who were not re-employed after the injury earned less before the injury (average of $13.09/hour and $552/week, converted to 1994 dollars) than did those who were re-employed after the injury ($17.82/hour and $821/week). The average hourly pay of the re-employed was only 4.9% lower after the injury than it was before the injury, but average hours worked fell from 42.4 to 35.8 hours per week and average weekly pay declined by 25%, while the average hourly and weekly pay of non-SCI former co-workers increased by 16%. In addition, the re-employed SCI individuals were 10% less likely to be covered by employer health insurance or by unions. Not surprisingly, white-collar employees before the injury were more likely to have post-injury employment, and many formerly blue-collar workers moved to white-collar jobs, so that the overall percentage of employees in white-collar jobs went from 49% to 83%.


EXTENT OF COMPUTER USE AND TRAINING

Computers have the potential to expand the work and nonwork activities of individuals with SCI's. How extensive is computer use and training? Two-thirds (66%) of the SCI individuals have ever used a computer, less than half (44%) of all SCI individuals first used a computer prior to the injury, and about half (46%) currently use a computer (as shown in Table 1, at end of document). The SCI individuals are less likely to be using computers than their non-SCI former co-workers. Computer usage is particularly high among quadriplegics, college graduates, and those who are younger.

Although less than one-third used a computer in pre-injury work, almost two-thirds (65%) of those re- employed currently use a computer at work (including 37% of those who had never used a computer prior to the injury). This compares to only a mild increase in computer use among the former co-workers without an SCI, from 42% to 47% (matching national trends from general population data). Post-injury computer training (from any source) has been received by less than one-fourth (22%) of the SCI individuals, with schools the most common provider. The lower rates of computer use and training among SCI individuals relative to the non-SCI sample can be traced to lower employment levels, since much computer use and training comes through the employer.


FACTORS AFFECTING EMPLOYMENT AND EARNINGS

What determines the employment and earnings of people with SCI's? Does computer use, which can help compensate for mobility limitations, have a particularly beneficial effect for individuals with SCI's? The analysis examines the roles played by demographic and injury characteristics, and computer use, on three outcomes of interest:

  • 1) whether the SCI individual has had post-injury employment or is currently employed;
  • 2) how long after the injury until paid employment is gained; and
  • 3) hourly and weekly earnings on the post-injury job.

Education is a powerful predictor of all three outcomes of interest. A majority of college graduates in the SCI sample are employed (61%) whereas only one-sixth (15%) of high school graduates are employed--a difference that holds up when controlling for other influences upon employment, and is much larger than the effect of education on employment rates in the general population. This is consistent with prior research indicating that education is a strong influence on the employment rate for people with SCI's. In addition, higher education appears to lead to a quicker return to work after an injury: the chance of re- employment in any given month is four times higher for college graduates than for those with no more than a high school degree, leading to a much shorter time until re-employment. For those who are employed, the strong effects of education on earnings (about 60-75% higher earnings for college graduates compared to those with no more than a high school degree) are similar to the effects found in the general population.

The ability to drive (which 70% in the sample can do) is a strong correlate of employment and post-injury time until employment, with very few non-drivers (5%) currently employed. Those with more severe injuries (quadriplegia and injuries where the spinal cord was completely severed) are less likely to be employed (with employment rates of 27% for quadriplegics and 35% for paraplegics), and take longer to find employment, but once employed receive pay that is comparable to those with less severe injuries. Women with SCI's are almost as likely as comparable men with SCI's to be employed, but their hourly pay is about 20% lower (consistent with the pattern in the general population), whereas blacks with SCI's are less likely than comparable whites to be employed, with 30% lower hourly pay (a greater black-white pay differential than in the general population).

The current employment rate for those with SCI's (30%) is less than half the rate that would be predicted for this group (76%) in the absence of an SCI, based on those with similar demographic characteristics in the general population. Disability income is strongly related to employment status (those receiving it are less likely to be employed), but it is difficult to disentangle the causality: low prospects for employability are often a criterion for receiving disability income, while the receipt of disability income often involves disincentives for employment due to reduced benefits if income is earned.


VALUE OF COMPUTER USE

Do computer skills make a difference? This question is addressed in two ways: first, by examining whether existing skills at the time of the injury help produce a quicker return to employment and better earnings recovery, and second, by examining the earnings premium from current computer use at work and comparing pay levels of computer users and non-users in the SCI sample and the general population. Having used a computer in pre-injury work is not significantly related to whether one has any post-injury employment, but is estimated to shorten the time until such employment is gained (reducing the expected time until employment by 40%). In addition, pre-injury computer use at work is associated with a substantially better recovery of earnings in the years following an injury.

Current computer use at work appears to significantly enhance the earnings power of people with SCI's and of the general population, even after controlling for the effects of education, experience, job seniority, union status, gender, and race. Table 2 shows the average hourly and weekly earnings we predict for a group of workers with the characteristics of SCI workers, separately by SCI status and computer use status. Among SCI workers, those who use a computer at work earn $873 per week, while those who do not use a computer at work earn just $376 per week--roughly a $500 earnings premium for those who use a computer at work. This is substantially larger than the $275 premium for computer use estimated for the general population. Indeed, SCI workers who use a computer at work earn essentially the same amount per week as non-SCI workers who use a computer at work, while among non-users of computers the pay of SCI workers is lower than that of non-SCI workers. Interestingly, those with SCI's who use a computer at work tend to work substantially more hours per week, and are more likely to hold full-time jobs, than are SCI workers who do not work with computers.

Computer skills can open up opportunities in high-paying occupations as well as improve prospects within occupations. Among SCI workers in a fixed set of occupations, computer users earn about 40% higher hourly earnings and 140% higher weekly earnings than non-users. The hypothesis that computers make a difference is also supported by examining the pay changes between pre- and post-injury jobs, where moving from a non-computer-using job to a computer-using job is found to be associated with similar significant increases in pay.


CONCLUSION

The results indicate that an SCI severely restricts workers' employment opportunities. Computer skills appear to be of particular benefit to people with SCI's in enhancing employment prospects and earnings. Despite their salutary effect, computer use and training are lower among those with an SCI than among former co-workers who were not injured. An important reason for the lower use of computers by the SCI population is that most people learn to use computers at work, and a very low percentage of individuals work after an SCI.

This evidence lends support to policies and programs that seek to increase access to computers and computer training for people with SCI's. While this study focuses on SCI (since the effects of computer skills can be most cleanly identified when focusing on a single traumatic disability that permits pre/post comparisons), the results suggest similar potential benefits for people with other kinds of disabilities (particularly for the other 300,000 wheelchair users under the age of 65). As computer use spreads, it is important to monitor whether people with disabilities are fully using computers to develop new productive abilities and employment opportunities.


REFERENCES

Berkowitz, Monroe, Carol Harvey, Carolyn Greene, Sven Wilson, "The Economic Consequences of Traumatic Spinal Cord Injury" (New York: Demos, 1992).

Stover, Samuel, Joel DeLisa, and Gale Whiteneck, "Spinal Cord Injury: Clinical Outcomes from the Model Systems" (Gaithersburg, MD: Aspen Publishers, 1995).

Trieschmann, Roberta, "Spinal Cord Injuries: Psychological, Social, and Vocational Rehabilitation" (New York: Demos, 1988).


For a copy of the full report on which this paper was based, "Disability, Employment, and Earnings in the Dawn of the Computer Age," contact Professor Kruse at the addresses or numbers above.

TABLE 1: Computer use and training

  Full SCI sample Those with no
pre-injury use
Paired sample
SCI
Paired sample
Non-SCI
Currently use computer at all 46% 27% 41% 54%
Currently use computer at school 9% 5% 4% 1%
Currently use computer at home 35% 21% 31% 32%
Currently use computer at work 19% 9% 20% 42%
If employed, use computer at work 65% 37% 61% 47%
If computer user, use adaptive
software or device
13% 15% 10%  
Current computer use if:        
Paraplegic 44% 21%    
Quadriplegic 51% 35%    
Female 43% 26% 36% 58%
Male 48% 28% 43% 51%
Age < 30 64% 54% 50% 62%
Age 30-39 50% 26% 56% 48%
Age 40+ 32% 17% 31% 53%
No high school degree 7% 6% 0% 13%
High school degree 26% 11% 21% 40%
Some college 54% 40% 38% 77%
College grad 85% 72% 69% 69%
Graduate degree 76% 67% 82% 90%
Received any training post-injury 22% 18% 16% 21%
If received training, provided by:        
Employer 23% 8% 27% 57%
School 61% 70% 36% 21%
Rehab center 9% 14% 9% 0%
Other 12% 14% 27% 21%

TABLE 2: Earnings of computer users and non-users

  SCI Non-SCI*
Adjusted hourly wage**    
Computer non-user $13.25 $14.11
Computer user 18.03 18.19
Percent premium for computer use 36% 29%
Adjusted weekly wage**    
Computer non-user $376 $584
Computer user 873 859
Percent premium for computer use 133% 47%

* Based on Current Population Survey data from U.S. Bureau of Labor Statistics.

** Earnings have been adjusted using regression to control for observable characteristics such as age, gender, marital status, and education. Reported figures reflects pay for someone with average characteristics of an SCI worker.