1 NBER WORKING PAPER SERIES IS AMERICAN PET HEALTH CARE (ALSO) UNIQUELY INEFFICIENT? Liran Einav Amy Finkelstein Atul Gupta Working Paper 22669 http://www.nber.org/papers/w22669 NATIONAL BUREAU OF ECONOMIC RESEARCH 1050 Massachusetts Avenue Cambridge, MA 02138 September 2016 We thank Nicolette Zarday for help with the pets claim data. The views expressed herein are those of the authors and do not necessarily reflect the views of the National Bureau of Economic Research. NBER working papers are circulated for discussion and comment purposes. They have not been peer-reviewed or been subject to the review by the NBER Board of Directors that accompanies official NBER publications. © 2016 by Liran Einav, Amy Finkelstein, and Atul Gupta. All rights reserved. Short sections of text, not to exceed two paragraphs, may be quoted without explicit permission provided that full credit, including © notice, is given to the source.

2 Is American Pet Health Care (Also) Uniquely Inefficient? Liran Einav, Amy Finkelstein, and Atul Gupta NBER Working Paper No. 22669 September 2016 JEL No. H51,I1,I13 ABSTRACT We document four similarities between American human healthcare and American pet care: (i) rapid growth in spending as a share of GDP over the last two decades; (ii) strong income- spending gradient;(iii) rapid growth in the employment of healthcare providers; and (iv) similar propensity for high spending at the end of life. We speculate about possible implications of these similar patterns in two sectors that share many common features but differ markedly in institutional features, such as the prevalence of insurance and of public sector involvement. Liran Einav Atul Gupta Stanford University Stanford University Department of Economics Department of Economics 579 Serra Mall 579 Serra Mall Stanford, CA 94305-6072 Stanford, CA 94305-6072 [email protected] and NBER [email protected] Amy Finkelstein Department of Economics, E52-442 MIT 77 Massachusetts Avenue Cambridge, MA 02139 and NBER [email protected]

3 Is American Pet Uniquely Inefficient? Health Care (Also) A IRAN E INAV , A MY F INKELSTEIN AND L TUL G UPTA * By , * Einav: Department of Economics, Stanford University , 579 Serra is not Naturally, the conventional wisdom -mail: [email protected] -6072 (e Mall, Stanford, CA 94305 ) and without its skeptics. An alternative school of NBER; Finkelstein: Department of Economics, Massachusetts Ins titute -mail: of Technology, 50 Memorial Drive, Cambridge, MA 02142 (e thought is that high and rising US healthcare [email protected] ) and NBER; Gupta : Department of Economics, ng is an optimal outcome given spendi - Stanford University, 579 Serra Mall, Stanford, CA 94305 -6072 (e mail: [email protected] thank ). We Nicolette Zarday for help with individual preferences. For example, Hall and . the pets claim data Jones (2007) argue that healthcare is a luxury It is well documented that the level and good (i.e. with an income elasticity above 1) growth of the US healthcare sector is high and calibrate a dynamic utility model under relative t o any other developed countr y, and which the observed rise in the US health s hare that this higher spending is not associated with of GDP is optimal. A related line of argument better health outcomes. Economists and emphasizes the dramatic technological frequently attribute these facts to policymakers progress in medicine and the value of life , idiosyncratic, institutional features of the US suggesting that high and rising US healthcare healthcare sector, focusing particularly on spending may be socially desirable (e.g. generous health insurance coverage that Muphy and Topel 2003, Cutler 2004). insulates consumers from the direct financial These divergent perspectives are intriguing, ion consequences of their healthcare consumpt and difficult to “resolve” with a single decisions, and public sector reimbursement and it may well be that convincing answer. Indeed, regulation that provides little incentive for a single answer does not exist, and the unique providers to engage in efficient production (e.g. spending patterns of the US healthcare system Weisbrod 1991; Fuchs 2014). Such features factors, some of result from a combination of have been suggested to be the cause of what which reflect specific institutional features of makes the American healthcare system, in the d some of which the American system, an words of Alan Garber and Jonathan Skinner reflect “deeper primitives ” concerning r and Skinner (Garbe “uniquely inefficient” individual preferences over health and 2008). healthcare or the nature of the supply -side of health care. Empirical p rogress on this

4 question is challenging in light of the fact that In the rest of t he paper we document four similarities between American human trying to explain the “uniqueness of the US comes down to the healthcare healthcare spending and American pet typically system” healthcare spending: (i) comparison of one single data point (the US rapid growth in healthcare system over the last few decades) to spending as a share of GDP over the last two (ii) similar -spending data points in other countries. decades ; a strong income gradient ; (iii) rapid growth in the employment In this paper, we offer a new data point by of healthcare providers ; and a similar simple facts about a different presenting some (iv) propensity for high spending at the end of life industry: the American healthcare industry. pet More details on the data, We show that many features of the American in pets and humans. y, pet health care sector are, qualitativel variable definitions, and analyses are presented remarkably similar to those of the American in the online appendix. this short We view the primary purpose of human health care sector . paper as bringing these facts into our collective Despite all the obvious caveats when consciousness to stimulate further discussion presenting human care spending and pet care and insights. In the concluding section, we spending in the same picture, the two industries share a common feature: the need to make offer some initial thoughts of our own. decisions and tradeoffs with respect to medical I. Patterns of Pet Care Spending Over Time spending that may potentially improve or ups and Across Income Gro extend life. Yet institutionally they are quite different: insurance for pet healthcare is much We use annual data from the Consumer 1 less common, and regulation (and public Expenditure Survey (CEX) from 1996 -2012 to sector involvement more broadly) less document patterns of spending on pets, and prevalent. The similarities we find in the compare it to three other spending categories: empirical patterns therefore point to deeper (human) healthcare, housing, and demand and that are also influencing primitives (Human) healthcare spending in entertainment. supply of health- related products. the CEX represents out of pocket spending by the household on health insurance premiums 1 ), while the American Pet Products Association -2016 industry -report Insurance rate appears to be less than one percent. The North reports on a national survey of pets owners, according to which there America Pet Health Insurance Association reports that 1.6 million pets are more than 160 million dogs and cats owned as pets in 2015 -16 -news/state 015 were insured in 2 - https://naphia.org/news/naphia ( (http://www.americanpetproducts.org/press_industrytrends.asp).

5 and healthcare. We choose housing and The growth of spending for each category is entertainment somewhat arbitrarily, as two presented in Figure 1: we normalize each 3 spending category by its 1996 level other normal goods, that are likely to correlate and s positively with income, within and acros the growth pattern in each category present households. The CEX measure of spending on over our observation period. While housing pets is composed primarily of two roughly ve been fairly and entertainment spending ha categories: spending on -sized sub- similarly flat over the 1996- 2012 period, healthcare “pet purchases and medical supplies ” and on spending has been steadily rising, with .” We group these together “veterinary services spending in 2012 being almost 50% high er than here, and show in the online appendix that in 1996. This rapid growth in healthcare if we restrict to just patterns are similar been widely spending has, of course veterinary services. documented and commented on previously. We annualize spending, so that our unit of The key observation from Figure 1 is that the year, and convert observation is a household- growth in spending on pets has followed -U spending to 2012 dollars (using the CPI healthcare spending remarkably closely, with price index). We limit our analysis to pet- 2012 spending being 60 % higher than spending who range owners ( of between 31 percent on pets in 1996. households in 1996 to 35 percent in 2012, with 1.8 a peak of 39 percent in 2010) by conditioning Healthcare Pets 1.6 Entertainment on household- years that report positive Housing 1.4 spending on pets. All analyses use the CEX sampling weights, which attempt to make it 1.2 repre sentative of the US population. Our final 1 sample covers 84,341 household- year 0.8 Annual spending (CPI adjusted) per household (relative to 1996) observations, which cover 57,346 unique 0.6 1996 2002 2001 2000 2005 2004 2003 1999 1998 2006 2007 2008 1997 2012 2011 2010 2009 2 households. Ye a r ARE C P ROWTH OF G 1 IGURE F ET PENDING S . 2 year level. Observations of the same household across calendar years The CEX conducts its interviews every quarter, with spending of are treated separately. -consecutive participating households typically observed for four 3 Spending levels across categories are naturally very different. quarters, which do not necessarily conform to calendar years. We aggregate quarters within a year, and then annualize to the calendar Housing spending per household in 1996 is $25,818 (in 2012 dollars), 5, entertainment is $7,744, and pets is $1,177. healthcare is $5,43

6 Figure plot -of -pocket spending per household in four s annual out highest income category (annual income spending categories. Spending is CPI adjusted and is normalized by the 1996 spending of each spending category. Sample include all greater than $70,000) spending between 114% households in the CEX with positive spending on pets. See text for more details. (for pets) to 2 ) more 59% (for entertainment households in the lowest income category. than ore how spending on each We also expl we find the spending patterns for Second, again category varies with income. To do so, we use healthcare by income to be similar to those of the same sample, and for each category ; indeed, we . This was not obvious a- priori pets compute the average annual spending by expected that health insurance would flatten income (using the categorical income brackets this relationship for human healthcare relative F igure 2 presents the the CEX). in available to pet health care, where insurance or other zing each spending estimate by results, normali redistributional policies are less common. the average household spending of the lowest income bracket ($20,000 and less) for the II. Growth of The Pet Care Sector corresponding spending category. data from the annual In this section we use 3.5 from 1996- County Business Patterns (CBP) Healthcare 3.0 Pets Entertainnment 2013 to document employment and Housing 2.5 establishment growth for veterinarians and 2.0 veterinarian -related services and compare it to 1.5 employment and establishment growth for 1.0 and physician physicians -related services. Relative annual spending (CPI adj.) per household Figure 3 shows the results. We show 0.5 20-30k 30-40k 40-50k 50-70k >70k <20k Annual HH income ($) employment in each sector and overall relative S PENDING BY I NCOME 2 C IGURE P ET ARE . F to its 1996 levels. Somewhat similarl y to the s the relationship between household income and annual Figure plot spending in four spending categories. Spending is CPI adjusted and is growth in spending (Figure 1), we see that the normalized by spending of the lowest income bracket for each spending category. Gray bars report the share of households in the significantly supply of physicians has grown lude all baseline sample in each income category. Sample inc households in the CEX with positive spending on pets except for 8% faster than employment growth in other sectors with missing income data. See the online appendix for more details. (but, interestingly, slower than the spending We make two observations from the results. growth). Yet, supply of veterinarians grew First, not surprisingly, all spending categories even faster: while the number of physicians in fairly st rong correlation between exhibit a 2013 was about 40% higher than that in 1996, income and spending, with households in the

7 e dogs, we the number veterinarians almost doubled over observation period. For thos the same period. The pattern of establishment obtained detailed information about their growth appear similar. claim -by-claim bills, and aggregated total spending as a function of the number of months 2 prior to death. In physician offices 1.8 a similar data extract for We then created In veterinarian offices US total employment 1.6 Medicare patients. Using data on beneficiaries ditional, fee -for -service Medicare, we in Tra 1.4 randomly selected 1 25 beneficiaries who were 1.2 Total employment (relative to 1996) diagnosed with lymphoma and died in 1 December (so , 2010, or 2011 of 2008, 2009 0.8 2009 2006 1997 1996 2002 2003 2004 2005 2007 2008 2000 2010 2011 2012 2013 2001 1999 1998 twelve months of claims data prior to death are Ye a r 1.4 observed). U sing detailed claim -level Physician-related establishments Veterinarian-related establishments 1.3 ion, we construct in parallel total informat US total establishments medical spending and used the claims data to 1.2 aggregate total spending as a function of the 1.1 number of months before death. Total establishments (relative to 1996) Figure 4 presents the main results. 1 Separately for the small sample of deceased 0.9 2012 2011 2010 2009 2001 2007 2006 2005 2004 2003 2002 1996 2000 1999 1998 1997 2008 2013 Ye a r dogs and the larger sample of deceased IGURE 3 ECTOR S F . G ROWTH OF T HE P ET C ARE Medicare beneficiaries, we normalize spending annual employment (top) and annual number of s Figure plot spending in the sample by the average monthly . It is sectors and for the US overall two the establishments (bottom) in County Business Patterns ( CBP ). based on data from the US census’ 10 to 12 months before death ( which is $414 for the average dog and $ for th e average 3,290 -of-Life Spending Patterns III. End Medicare beneficiary), which we define (with all the obvious caveats) as a “regular month.” We obtained a small extract of billing data life -of- As one can see, there is a distinct end from a single pet hospital in California. The . spike in spending for both populations hospital provided us with data on a randomly Relative spending in the last month before selected sample of 44 dogs who were treated death is large of a “ regular ” month for : 2.18 We for lymphoma between 2011 and 2014. Medicare beneficiaries and 3.48 of a “relative focus on 23 of these dogs who died within our

8 month” for dogs. The horizon over which IV. Discussion , spending spikes is slightly dogs different; for We presented several descriptive patterns there is little “excess” spending before the last about the pet health care industry in the United month (e.g., two months before death spending States, which overall appear to be qualitatively is only 30% higher than a regular month), while and -documented similar to parallel well for Medicare beneficiaries there seems to be an health patterns of the US (human) discussed -4 months elevated level of spending already 3 care sector. prior to death. All the obvious and appropriate caveats 4 Mean (healthcare) Mean (dogs) 3.5 associated with the comparison of human Median (healthcare) Median (dogs) 3 hat health care and pet care notwithstanding, w 2.5 drew us to the study of pet health care is the 2 many similarities in the nature of the consumer 1.5 Relative monthly spending choice problem, juxtaposed with sharp 1 differences in the institutional environment in 0.5 0 which the choice is made. The two industries 10 8 7 6 5 4 3 2 9 1 11 12 Months before death . From demand share many similarities IGURE F PENDING IFE S -L - ATTERNS P ND E . 4 OF perspective, treatment decisions are triggered a for Figure plot s monthly spending for the 12 months prior to death -diagnosed dogs and a larger sample of small sample of 23 lymphoma by health episodes that are often difficult to onthly spending -diagnosed Medicare beneficiaries. M lymphoma 125 is normalized by average monthly spending in the 10 -12 months pri or to death. forecast, they are channeled by expert intermediaries who may not fully internalize Of course, although we find the patterns the financial cost associated with treatment, resting, it is important to note that unlike inte involve and they often emotional and financial which the rest of the analysis in this paper – tradeoffs supply perspective, the nature . From the data uses standard, national data sources – of technological progress is similar , and on end of life spending for dogs with provision is channeled by lengthy education lymphoma relies on an small sample of dogs and training and the requirement for h likely whic from one specific pet hospital occupational licensing . But, in contrast to these draws customers who are significantly richer between pet healthcare and human similarities than the average dog owner. care in the nature of the consumer’s health choice, the institutional environment is very different for pet healthcare.

9 Most notably, insurance is much less case pet health care – suggests the potential , or common in pet care, and regulation importance of further work seeking to understand preferences over health – in government involvement more broadly, is not addition to the traditional study of insurance, as prevalent. The fact that despite these differences – in understanding often mentioned as potential incentives, and institutions – explanations for the large and rapidly growing the US healthcare spending and treatment some pet health healthcare sector in the US – patterns. care patterns appear qualitatively quite similar REFERENCES to the analogous human health care pattern, give us us as noteworthy. It should strikes Chandra, Amitabh, Amy Finkelstein, Adam pause before attributing the large and rising Sacarny, and Chad Syverson. 2016 to the healthcare costs in the US solely “Healthcare Exceptionalism? Performance prevalence of insurance and government and Allocation in the U.S. Healthcare 4 involvement. . The similar growth patterns in Sector.” c Review , American Economi may also suggest US human and pet healthcare forthcoming. t technological change in human healthcare tha Your Money or Your Life: Cutler, David. 2004. may have spillover effects on related sectors, Strong Medicine for America’s Health Care including perhaps pet healthcare or human care System (New York, NY: Oxford University in other countries. Press, 2004) Of course, much more work is needed to Why Do Other Rich Fuchs, Victor R. 2014. “ explore this further. But at some broad level , Nations Spend So Much Less on cal similarities between pet and these empiri The Atlantic HealthCare?” , July 23, 2014. human health care follow the spirit of Chandra Garber, Alan M., and Jonathan Skinner. 2008. who suggest that the US et al. (forthcoming) “Is American Health Care Uniquely healthcare sector may not be as unique as often Journal of Economic Inefficient?” is claimed, and may benefit from economic 22(4), 27-50. Perspectives insights gleaned from studying other industries. Hall, Robert, and Charles I. Jones. 2007. “The Here, our study of another industry – in this Value of Life and the Rise in Health 4 The spirit of Tu and May (2007), who find limited shopping -pay -related self behavior by consumers in the context of health markets, is quite similar.

10 Spending.” Quarterly Journal of Economics 122(1), 39 –72. Murphy, Kevin M., and Robert Topel. 2003. .” “The Economic Value of Medical Research Measuring the Gains from Medical in Research: An Economic Approach , Kevin M. Murphy and Robert Topel, eds. (Chicago: Uni versity of Chicago Press). Tu, Ha T., and Jessica H. May. 2007. “Self -Pay Markets In Health Care: Consumer Nirvana Health Affairs 26(2), Or Caveat Emptor?” w217-w226. Weisbrod, Burton A. 1991. “The Health Care Quadrilemma: An Essay on Technological Change, Insurance, Quality of Care, and Cost Containment.” Journal of Economic Literature XXIX, 523-552.

11 Online Appendix ) Uniquely Inefficient?” “Is American Pet Health Care (Also Liran Einav, Amy Finkelstein, and Atul Gupta Consumer Expenditure Survey Data (Figures 1 and 2) Figures 1 and 2 use data from the Consumer Expenditure Survey (CEX) from 1996-2012. Prior to 1996 the format of the CEX files differs substantially; 2012 was the latest available year at the time we began this project. The CEX conducts interviews with households over 5 quarters – the first quarter is a baseline interview, while the remaining 4 quarters record expenditures (“consumption”) each quarter. These four quarters do not necessarily conform to a calendar year. We aggregate quarters within a calendar year. Observations of the same household across calendar years are treated separately. The unit of observation is therefore a household- year. When households do not appear in the survey for all four quarters of a calendar year, we use the available quarters and annualize the total to obtain annual expenditure levels. We analyze data on four categories of expenditures: pet care, (human) health care, housing, and entertainment. Human health care includes the household’s (out of pocket ) spending on health insurance premiums, physician, and hospital services. Housing includes spending on mortgages, rent, property taxes, and maintenance costs. Entertainment includes spending on recreational activities and equipment, and television subscriptions and equipment. Our measure of pet care is based on three expenditure categories: ( i) pet purchase and medical supplies; ( ii) Veterinary services ; and ( iii) pet services except Veterinary services. These sub - categories account , respectively, for 50, 38, and 12 percent of total pet care spending. Of the two main categories, v eterinary services are naturally part of pet health care, while some, but not all of the “pet purchase and medical supplies” category is too. It is unclear what is covered by the smaller “pet services except veterinary services” category. In our main analysis we aggregate all three sub- categories . In Appendix Figures 1 and 2 below we replicate Figures 1 and 2 in the main es” category text limiting spending on pets to just the “Veterinary servic ; the results are qualitatively quite similar (the sample remains the same; a pet owner is still defined as a household with positive spending on pets, regardless of whether the spending is on veterinary services). We start with the full CEX sample in every quarter, omitting only households (less than one percent) where the head of the household is younger than 18 or older than 90. This results in a sample of 240,390 household- years. We further limit all analyses to ho usehold- years with positive expenditure on pets, which is our proxy for pet ownership. About 35 percent of household- years are included as “pet owners ,” so that our final sample size is 84,341 household- years, cover ing 57,346 unique households. Appendix Ta ble 1 below compares demographics and expenditure on the main categories between our baseline sample and the entire CEX sample. 1

12 years by the When we analyze expenditure by income in Figure 2, we categorize household- household income bucket they report. Hou sehold income in the CEX is categorized into <20,000, 20,000 – 29,999, 30,000 – 39,999, 40,000 – 49,999, 50,000 – 69,999, >=70,000. The income category is missing for 8% of the observations. 1.8 3.5 Healthcare Vets Healthcare 1.6 3.0 Entertainment Vets Entertainnment Housing Housing 1.4 2.5 1.2 2.0 1 1.5 1.0 0.8 Relative annual spending (CPI adj.) per household Annual spending (CPI adjusted) per household (relative to 1996) 0.5 0.6 50-70k 30-40k >70k 40-50k 20-30k <20k 2000 2001 2002 2003 2004 1996 2006 2007 2008 2009 2010 2011 2012 1997 1998 1999 2005 Annual HH income ($) Ye a r gure 1 Appendix Fi gure 2 Appendix Fi Expenditure growth (vets) Spending by income (vets) Entire CEX sample HHs w/ positive pet spending HH and head of HH Demographics: White 0.92 0.83 Male 0.50 0.51 0.01 Age 18-24 0.02 Age 25-64 0.76 0.84 Age 65-90 0.15 0.21 Family size 2.73 2.52 College degree 0.42 0.37 HH income >= 50k 0.49 0.37 Expenditure categories: Pets, medical supplies (2012$) 369 130 Veterinarian services (2012$) 283 100 Pet Services (except Vet) (2012$) 93 33 2,956 3,532 Health care (2012$) Entertainment (2012$) 3,470 2,587 10,185 11,626 Housing (2012$) 84,341 Observations (HH-year) 240,390 Appendix Table 1 (1st col) and the entire CEX sample (2nd col) Summary statistics for our baseline sample 2

13 US County Business Patterns (Figure 3) We use annual data from 1996-2013 from the County Business Patterns (CBP) published by the US Census. We aggregated employment data across standard industry classification codes (the We analyze total North American Industry Classification System, or NAICS ) to the sector level. veterinarians . and employment as well as employment in two specific sectors: physicians Physician employment is defined as employment in hospitals, physician offices, dentist offices , and all other health care professional offices. Veterinarian employment is defined as employment at a veterinary office; both veterinary and physician employment measures will therefore include support staff in those offices. We sum employment across counties to arrive at national, annual totals. defines an establishment as “a single physical location at which business is conducted The CBP or services or industrial operations are performed.” An establishment is not necessarily equivalent to a company or enterprise, which may consist of one or more establishments. A single -unit company owns or operates only one establishment. A multi-unit company owns or operates two or more establishments. The series excludes data on self -employed individuals, employees of private households, railroad employees, agricultural production employees, and most government employees. End of life care (Figure 4) The data on end of life care for dogs come from a large veterinarian hospital in California. Using information on billed drugs, they identified animals who were treated for lymphoma between 2011 and 2014 (the period for which electronic billing data was available). We limit our analysis to dogs, which account for 80 percent of their patients. We received data for a random sample of 300 dogs who had received a biopsy, which is a diagnostic test for lymphoma, among other , and we therefore things. Of these 300 dogs, 44 were identified as having received chemotherapy code them as having lymphoma. Of those 44, we have 23 dogs who died during the period of our data and for whom we observe billing claims for at least 12 months before death. Our analysis of end of life spending pertains to these 23 dead dogs. For those dogs, we obtained detailed n about their claim- by-claim bills, and use this to create a monthly measure of total informatio spending for each of the 12 months prior to death. The data on end of life care for humans is based on claims data from Traditional Medicare. We selected a random sample of 1 25 patients who were diagnosed with Lymphoma lates t by the penultimate year of life (i.e. if the patient died in 2008, she was diagnosed with Lymphoma by 2007 or earlier), and who died in December of a year between 2008 and 2011. Patients were identified as having Lymphoma based on ICD9 diagnoses codes in the 200xx (Lymphosarcoma and reticulosarcoma) or 202xx (Other malignant neoplasms of lymphoid and histiocytic tissue ) families. These two groups are used by the AHRQ Condit ion Classification System (CCS) to define Lymhoma. For these 125 patient s, we measure monthly total Medicare spending in each of the 12 months prior to death. In both cases (dogs and Medicare), we observe claims for the 3

14 last 12 months of life. The spending values span several different years and hence the monthly values are inflation adjusted to be in 2012 dollars using the CPI- U series. 4

Related documents