In September 2012, The Los Angeles Times ran a story suggesting that the Affordable Care Act (ACA) would usher in a new age of health care consumerism powered by on-line tools, cost calculators, and armies of engaged consumers. In the intervening years, what has been repeatedly pronounced is that we want consumers to “have skin in the game” and for them to “buy value” by considering both price and quality when purchasing health care services.

Our respective employers, the Health Care Cost Institute (HCCI) and Consumers Union, endorse efforts to increase consumers’ effective use of price and quality information and to improve the efficiency of health care markets. However, the excessive expectations regarding the promise of health care consumerism need to be recognized before we design systems that depend on patients’ being savvy shoppers.

How Much Can Consumers Actually Shop For Health Care Services?

Less than 7 percent of our health care spending is actually paid by consumers on shoppable services. That means that for the vast majority of health care spending, consumer-targeted activities—such as providing incentives or information in an effort to direct their spending—are unlikely to improve the value we get for our health care dollars.

Instead, payers, employers, and other stakeholders need to consider a broader approach, including engaging providers directly. While knowledge about price and quality is important, and is information that consumers are entitled to, we should be realistic about the ability of this information in the hands of consumers to drive changes in the market.

This blog summarizes new findings from HCCI estimating spending on medical services that are shoppable, the proportion paid by consumers, and the level of price variation by broad categories of services. Before our discussion of shoppable services, it should be noted that:

  1. Roughly 25 percent of the commercially insured population does not have a claim in a given year; engaging these individuals as health care consumers is not immediately productive.
  2. Approximately 50 percent of health care dollars are spent on 5 percent of the population: Engaging these sick individuals as shoppers will be difficult, and suggesting that they actually shop could be inconsistent with other goals such as getting severely unhealthy individuals into medical homes, Accountable Care Organizations, or integrated delivery systems that can coordinate their care.
  3. While some health care is shoppable, our estimates represent an upper bound on the share of dollars spent on such services. For a variety of reasons, some services classified as shoppable cannot always be planned for in advance. For instance, though we characterize all knee and hip replacements as shoppable, the hip replacement after a fall is likely to be less shoppable than a scheduled surgery. Further, the absence of price variation among providers for shoppable services will result in little or no value in shopping. A prime example of this would be the presence of only one provider in any market for a given service (such as only one hospital in a community).

HCCI estimates that in 2011, total spending on behalf of the national employer-sponsored insurance (ESI) population (younger than age 65) was $524.2 billion. Following the methodology developed by Chapin White and Megan Eguchi, a shoppable service must typically be scheduled in advance, there must be more than one provider in a market that can perform the service, and price data have to be available for the different providers. Prescription drugs and devices are considered non-shoppable services in this analysis. Replicating the White and Eguchi methodology using the HCCI dataset, weighted to reflect the national population, we found that, at most, 42.5 percent of total spending was on medical services that can be considered shoppable (Figure 1).


Though 42.5 percent of the $524.2 billion spent per year on shoppable services is still a very sizable number ($222.9 billion), consumers pay out of pocket roughly 15 percent of this amount or about $37.7 billion — 7 percent of total spending.

Several benefit design features influence what a consumer pays out of pocket. The HCCI data show that consumers hoping to use price shopping as a way to lower their coinsurance and deductible payments will be able to influence, at most, an average of 65 percent of their out-of-pocket payments for the year. Copayments—the other big portion of out-of-pocket spending—are generally set amounts and will not vary based on consumer shopping.

Having identified an upper bound on the share of health care services that are potentially shoppable, an important question is how much price variation exists within shoppable services. In other words, what are the potential savings from shopping for services? Without price variation, there is little to gain from shopping.

How Much Is Saved By Shopping?

Our analysis of the HCCI dataset indicates that, in general, there is more price variation in inpatient services than in outpatient services. In this data, price is the amount paid on the claim. Furthermore, outpatient services tend to be lower-priced — suggesting that consumers have less to gain from shopping for outpatient services.

Interestingly, prices vary more for the non-shoppable inpatient services than for the shoppable ones. This could be because non-shoppable services have more inelastic demand due to the nature of the services.

Knee and hip replacements are inpatient, shoppable services that are a common target for efforts to steer consumers to the highest-value providers. Nationally, we find that knee and hip replacement admissions had less price variation compared to either inpatient shoppable (excluding knee and hip replacements) or inpatient non-shoppable admissions, as illustrated by a lower coefficient of variation. Table 1 illustrates general price variation across broad categories of services. However, this result is not to suggest that shopping for knee and hip replacements never makes sense. In Palm Bay, Florida, a knee replacement costs $16,822 more than the same surgery 180 miles away in Miami, Florida.

Table 1. Price Variation In Inpatient Services, For The National Population


Not surprisingly, geographic variation also suggests that shopping in some markets is more productive than in others. For example, we find much greater price variation across all three inpatient service categories in Kentucky, Texas, and Georgia than we find in either Montana or Hawaii (Table 2).

Table 2. Price Variation In Inpatient Services, For States With Most And Least Variation


Returning to the Palm Bay, Florida example above, a copay or deductible does little to entice a consumer to shop. Assuming an insured individual has satisfied his or her deductible and has a 20 percent coinsurance responsibility, that consumer’s savings from having the surgery in Miami would be $3,364—less the cost of getting to Miami—and not including any additional psychic or other costs of having the surgery far from home. Though this amount may be sufficient for some consumers to shop, the savings to the payer are far more significant: $13,458.

While consumers are entitled to usable shopping data on both price and quality for any medical service they could conceivably plan for in advance, we need to be realistic about the power of consumer shopping to rein in excess health care pricing when selecting medical services. Instead, we need to accompany consumer shopping tools with interventions that bypass the consumer. On the quality front, providers and payers will need to be engaged to ensure that the services provided meet an acceptable threshold. Efforts to reduce health care costs need to directly engage providers, who are setting the prices, and payers (whether insurers or employers), because they are in the best position to exert downward market pressure on prices.