The Veiled Economics of Employee Cost Sharing.

This year, once again,millions of people in the United States who get health insurance through their employers received the unwelcome news that cost sharing would increase. Harvard University, where both of us work and get our health insurance, increased costsharing for its employees, raising a hue and cry from faculty.1 There were charges that the changes were regressive and particularly harmful for lower-wage employees.1 The critiques implicitly presumed that it is possible to have high wages, lower premiums, and no cost sharing. But this presumption misses the fundamental economic connections between wages, premiums, and cost sharing. Cost sharing has certainly increased, from copayments forphysicianofficevisitsandprescriptiondrugs to deductibles; the fractionofworkers inaplanwithat least a $1000 deductible for coverage of a single person increased from 10% in 2006 to 41% in 2014.2 Higher cost sharing feels likeadecreaseboth in thegenerosityofcoverageand incompensation. It seemsparticularlyunfair to lower-wageworkerswho face the samedeductibles and copayments as their higher-paid counterparts and who may be discouraged from seeking needed care. But increases in cost sharing arenot necessarily regressivenor necessarily associatedwith lower compensation. The reality of who actually pays for health insurance drives the different impacts of changes in insuranceplanson low-wageandhigh-wageemployees.Despite the hand-wringing over increases in employee premium contributions, the employee share of premiumshas stayedbetween27%and29%for the last 2decades, although the dollar amounts have increased because total premiumshave increased. Thepremium for a family policy more than doubled from approximately $8000 in 2002 to $16 800 in 2014.2 This is far from transparent to employees, most of whom do not see theiremployer’s shareof thepremium.More important— but evenmore opaque—is the fact that employees ultimately pay not only their share of premiums but their employer's share aswell.3 This is drivenby theeconomicsof labormarkets.Employersare largely indifferentbetween paying an employee $40 000 in wages and $20 000 inbenefits andpaying$50 000 inwages and $10 000 in benefits—in both cases, total compensation is $60 000.When the cost of health insurance increases, less money is left available for wages. This “wage-fringe” trade-off is well documented and applies to nonprofit and for-profit employers alike. Increases in health insurance premiums do not get absorbed by an unlimited reservoir of prof its or endowments—they are paid for by employees taking home smaller paychecks.4,5 The trade-off does not occur instantaneously for each individual, however. So increases in premiums are muchmore visible and salient than their effect on take-home pay. The trade-off betweenwages and fringebenefits is central tounderstanding thedistributional effects of increases in health care costs. Employers provide a similarmenuof insurance options toworkerswith different wages and salaries. Health insurance premiums represent a much larger share of compensation for a family taking home $40 000 than for a family that makes $150 000—and a premium increase of $1000 takes a much bigger percentage bite out of take-home pay for the lower-income family. A lower-income family might prefer tohave less generoushealth insurance andmore compensation, so that more money was available for rent, gas, and other priorities. Sowhy do they have this compensation package? Akey reason thatemployersprovidea similarmenu of insurance options, regardless of an employee’s income, is that the tax code in the United States favors health insurancebenefits relativetowagesas longasemployersoffer theirhigh-and low-wageworkers thesame plans. This tax preference fosters compensation packagesthatareskewedtowardhealth insuranceratherthan wages. The skewinghas2 insidious effects: it is both regressive and inefficient. The tax preference for health insurance is regressive because it gives a greater tax benefit to higherincomeworkers: an employee in the 40%marginal tax bracketwith a $10 000 tax-free policy saves $4000 in taxes avoided, whereas an employee in the 15% tax bracket saves only $1500. Higher-income workers are alsomore likely to have jobs that offer expensive insurance plans. As a result, lower-wage workers have slow or nonexistent wage growth because of the growing share of their compensation devoted to health insurance instead of wages, and their insurance plans cater more to thepreferencesofhigher-wageworkers than to theirs. Remedying this regressiveaspectof the tax code is one of themotivations for the “Cadillac tax”: starting in 2018, health insurers have to pay a tax on employer health insurance plans with premiums greater than $10 200 for individuals or $27 500 for families.6 These dollar amounts increase only as quickly as inflation, so if health insurance premiums increase more quickly, moreandmoreplanswill besubject to the taxover time. TheCadillac tax provides amotivation for employers to slow premium growth. Another reasonto reduce the taxsubsidy forexpensiveemployer-sponsoredhealth insurance isthatthesubsidies encourage the proliferation of plans withminimal cost sharing,which in turnencourages the inefficientuse ofmedicalcare.Atfirstblush, itmightseemthatcostsharing is just awayofdividingupwhetheremployersor employeespay thebills, but decadesof evidence show that lower cost sharing leads patients to consumemore care of limited health value—such as unnecessary tests—and that this consumption leads to higher health insurance VIEWPOINT