On a key health-care issue, Trump and Gillibrand made opposite claims within days of each other. To hear Trump tell it, prescription drug prices have fallen for the first time in 51 years. According to Gillibrand, a candidate for the Democratic presidential nomination, they haven’t fallen.
The White House and Trump’s campaign pointed to the consumer price index for prescription drugs, which has seen annualized declines in six of the last seven months. Gillibrand’s campaign said that the CPI for prescription drugs is flawed and that prices haven’t fallen according to numerous independent studies.
Neither of them is entirely right, and both pick and choose what best supports their case. Let’s dig in.
Trump was referring to the consumer price index for prescription drugs, which is updated monthly by the Bureau of Labor Statistics. It measures the “price change of drugs purchased with a prescription at a retail, mail order, or Internet pharmacy,” according to BLS.
“The tracked price is the total reimbursement to the retailer from the patient and all eligible payers for a single prescription,” BLS says. “Payers are any entities who reimburse health care providers for the cost of medical services and/or goods. Eligible payer types for prescription drugs are: patient self-pay (cash), commercial or private insurance, and Medicare Part D.”
The White House and Trump’s campaign said the prescription-drug index declined 0.6 percent in 2018. That was the first calendar-year decline since 1972, making it a 46-year record, not 51 years as Trump tweeted.
As often happens with data, the picture changes depending where the baseline is set. It’s somewhat arbitrary to measure calendar years, as Trump did. With monthly data, it’s possible to measure any 12-month period and not just calendar years.
For example, in the 12-month period ended July 2013, the CPI for prescription drugs declined 0.2 percent. Viewing it that way, Trump’s record shrinks to 5½ years.
Trump’s campaign pointed out that a pattern of falling drug prices is starting to emerge in the CPI. “In six of the last seven months, the CPI for prescription drugs has gone down when measured on a year-over-year basis,” a Trump campaign spokesman said. For the year ended June 30, the index declined 2 percent. That’s the largest decline on record for any 12-month period in the BLS data, which goes back to 1970.
Experts we consulted said the CPI uses a flawed methodology to measure prescription drug costs, an argument echoed by the Gillibrand campaign.
“Focusing narrowly on the CPI, the president’s claim strikes me as either true or close to true,” said Matthew Fiedler, a health-care expert at the Brookings Institution who was an economist in the Obama White House.
“There is, however, a major caveat: The CPI is a highly imperfect measure of drug price trends because it does not capture the rebates that drug manufacturers pay to insurers and pharmacy benefit managers. Thus, the CPI essentially measures the list prices of prescription drugs, not true net prices. Rebates are frequently large, so trends in net prices can differ materially from trends in list prices.”
What happens when using a different measurement? The price of generic alternatives seems to be declining, but the same does not appear true for branded drugs.
“I tend to look at the producer price index because it is closer to what manufacturers are selling drugs for,” said Richard G. Frank, a professor of health economics at Harvard Medical School. “The recent pharmaceutical PPI is pretty flat for the last couple of months and up a bit since December 796.4 to 810.0 in May (1.7 percent). Market watchers like Drug Channels and IQVIA all suggest that through the end of 2018 that net price growth (after rebates) has slowed but were still positive, at about 1.5 percent. Gross prices (before rebates are accounted for) grew at an annual rate of 5.7 percent.
“Of course if you focus on generic prices only, that price index has declined. Bottom line, I think the evidence points to slowed growth not negative growth.”
A January 2018 study published by the National Bureau of Economic Research raised concerns about the adequacy of both the CPI and PPI for prescription drugs. From 2005 to 2016, the market share of generic alternatives grew from 50 percent to 85 percent, which could be skewing the data, the researchers said. An industry shift toward electronic transactions and third-party payers also could be skewing the data, making it more reliant on cash transactions at pharmacies, the NBER study concluded.
A Gillibrand spokesman said “CPI is problematic” and sent us fact checks from the Associated Press, PolitiFact and Kaiser Health News. All of them deemed similar claims from Trump and his administration to be false or mostly false. The Gillibrand spokesman pointed out that those fact checks and other recent reports, including one in The Washington Post, mention studies showing that pharmaceutical drug costs have not fallen.
- The AP reported: “An analysis by Altarum, a nonprofit research and consulting firm, found that in 2018, spending on prescription drugs was one of the main factors behind a 4.5% increase in U.S. health spending. Spending on prescription drugs grew much faster than in 2017, according to the study.”
- CBS: “About 41 drugs have boosted their prices by more than 100%, including one version of the antidepressant fluoxetine — also known as Prozac — whose cost has surged 879%, Rx Savings Solutions said.”
- The Post’s Health 202: “List prices for branded drugs grew 3.3 percent in this year’s first quarter, compared with 6.3 percent in the first quarter of 2018, according to SSR Health pharmaceutical analysts.”
- Politico: “Prescription drug prices jumped 10.5 percent over the past six months, more slowly than during the same period last year but still four times faster than inflation, despite increasing pressure on drugmakers from the Trump administration and Congress. However, the top drug brands in the U.S. saw prices rise by an average of only 3.1 percent over the past six months according to Bernstein analysts, who calculated the average based on which brands — usually the top-sellers — they cover. The data showing an overall leap in generic and brand-name prices was preliminary and may not give a clear picture of trends in the industry.”
In September, the AP reported that in the first seven months of 2018, “there were 4,412 brand-name drug price increases and 46 price cuts, a ratio of 96 to 1.”
A White House official said that the CPI was a reliable metric and that the AP did not include generic drugs in its analysis.
“The CPI measures transaction prices — meaning the total between the out-of-pocket payment and what is paid by third parties — which is a more relevant measure of price (particularly for those with insurance) that includes negotiated price concessions and discounts,” the official said.
The White House and Trump campaign both argued that the Food and Drug Administration’s efforts to speed up approvals for generics has pushed down consumers’ costs.
The Trump administration has made it less expensive for companies to apply for generic approvals. The FDA says it set a record for generic approvals in fiscal year 2018, at 781, breaking the record of 763 set in the previous fiscal year. (We previously gave Two Pinocchios to the president for claiming credit over $26 billion in consumer savings from generic drugs.)
“The AP only looked at branded drugs and measured changes in list prices, which is not the ultimate price to the patient,” the White House official said. “On the other hand, CPI measures all drugs — i.e., includes generics — which are 90 percent of the market volume. Those prices have been going down thanks to FDA’s deregulatory efforts and its speeding up of approvals, which has also given branded prices more competition until they go off patent.”
Fiedler said: “One thing to know, however, is that while generics are about 90 percent of prescriptions, they are only ballpark one-quarter of total spending since prices for generics are so much lower than prices for branded drugs.” He also questioned how much the FDA approval process for generics has contributed to the falling CPI: “Growth in the prescription drug CPI tends to be volatile and can change for many reasons unrelated to policy. Moreover, the main policy change the administration has highlighted in this context is steps it has taken to make it easier for generic drugs to come to market. I have not seen persuasive evidence on how those changes affected drug prices, and I am doubtful that any effects on prices would be large enough to explain much of recent trends.”
Trump dropped one major plank in his plan to reduce drug prices last week. His drug rebate rule would have ended a widespread practice in which pharmaceutical companies give rebates to insurance middlemen in government programs such as Medicare. The idea was to channel that money to consumers instead.
Days before Trump’s announcement, a federal judge blocked a new rule proposed by the Trump administration that would have required drugmakers to include the list prices of their medicines in television ads. That was another big component of the president’s plan to lower drug prices.
The Bottom Line
Trump’s claim to a 51-year record is just wrong. The CPI for prescription drugs in 2018 fell for the first time in 46 years. And that’s only when focusing on calendar years, which is somewhat arbitrary. The president’s record shrinks to five-and-a-half years when looking at the most recent 12-month decline, for the year ended July 2013.
There’s a more salient criticism here: Experts say the CPI for prescription drugs fails to account for rebates, which can be substantial, and may be giving a skewed picture because of recent market shifts toward generics and electronic payments by third parties. A range of studies we found shows that drug prices have not declined, especially when it comes to branded drugs. So Gillibrand has an amply supported case to make.
On the other hand, to see the CPI for prescription drugs decline year-over-year for six of the last seven months is nothing to sneeze at. The 2 percent decline for the year ended June 30 is the largest on record for any 12-month period in the BLS data, which goes back to 1970. Flawed though the index may be, Trump is signaling a real pattern in the most relevant government statistic.
Send us facts to check by filling out this form