Special Report
US services spending collapsed during the COVID-19 pandemic, and remains significantly below the level that would have prevailed had the pandemic not occurred. This raises the question of whether services consumption will ever return to “normal.” In this report, we address this question by examining the weakest components of services spending, with an eye towards any evidence indicating that this weakness is permanent. A category analysis of services spending highlights that the spending gap currently exists due to a combination of work-from-home trends and evidence of lasting aversion to COVID-19. The latter is unlikely to be permanent, and the former will be partially or fully offset by a permanent increase in substitutable goods spending. In a non-recessionary scenario, our analysis suggests that the US services spending gap will continue to close, which will provide support for overall consumption as goods spending slows in response to weak real wage growth and higher interest rates. The COVID-19 pandemic has been enormously disruptive, socially as well as economically. In the US, a massive shift from services to goods spending represents one of the most significant economic disruptions caused by the pandemic, which persists even today. Chart II-1The Pandemic Caused An Extreme Shift In Spending From Services To Goods Chart II-1 presents our best estimate of the real goods and services spending gaps relative to potential GDP, which illustrates how extreme the shift from services to goods has been. The real goods spending gap exploded during the pandemic to a level that had not been seen since the early-1950s, and services spending collapsed in an unprecedented fashion and remains at a level that is lower than at any other point over the past seven decades (aside from the worst of the pandemic itself). Chart II-2 highlights that the overall output and household consumption gaps have not yet turned positive, despite an extremely strong labor market. This underscores that weak services spending is playing a role in depressing consumption, and thus overall economic activity. Chart II-2Weak Services Spending Is Playing A Role In Depressing Consumption This persistent weakness in services spending raises the question of whether services consumption will ever return to “normal,” defined as the level of spending that would have likely prevailed had the pandemic never occurred. In this report we address this question by examining the weakest components of services spending, with an eye towards any evidence indicating that this weakness is permanent. We conclude that the services spending gap currently exists due to a combination of WFH trends and evidence of lasting aversion to COVID-19. While the effect of the former may be permanent, we do not believe that the effect of the latter will be. And, in cases where certain categories of services spending are likely to be permanently lower, at least some of this decline in spending is likely to be partially or fully offset by a permanent increase in substitutable goods spending. In a non-recessionary scenario, our analysis suggests that the US services spending gap will continue to close, which will provide support for overall consumption as goods spending slows in response to weak real wage growth and higher interest rates. The Pandemic, Remote Work, And Services Spending During the very early phase of the pandemic, COVID-19 was spreading rapidly in industrialized economies. Following recommended or mandatory stay at home orders from governments in many countries, most office-based businesses rapidly shifted to work-from-home (WFH) arrangements as an emergency response. This, in conjunction with forced closures of “close contact” businesses such as restaurants, entertainment, and travel caused US services spending to collapse. However, by the summer of 2021, many of these pandemic control measures had been significantly eased or lifted in the US. In addition, several national US surveys found many office workers preferred the flexibility afforded by WFH arrangements. Many employers, correspondingly, found that the productivity of their employees did not suffer while working from home, or that it even improved. These findings led many in the business community to conclude that WFH policies are not, in fact, emergency measures that will ultimately be reversed and instead reflect the “new normal” for work. While this “new normal” is still in the process of being defined, it seems fairly clear that some form of hybrid work arrangements will be permanent for many businesses. Chart II-3 presents the Kastle Systems Back to Work Barometer, which reflects keycard swipes in office buildings in the top 10 US cities. The chart highlights that urban office building activity has recovered to less than half of its pre-pandemic level, and that there has been no evidence of a continued uptrend over the past 3 months. Chart II-4 reinforces this point by highlighting that public transit use in major US cities has lagged the recovery in air travel, and also has not substantially changed over the past few months. Chart II-3Urban Office Building Activity Has Recovered To Less Than Half Of Its Pre-Pandemic Level Chart II-4Urban Public Transit Use Has Lagged The Recovery In Air Travel This underscores that investors have a basis to question whether at least some US services spending may be permanently impaired by the pandemic, as was the case for overall output for several years following the 2008/2009 global financial crisis. To answer this question, we present a detailed review of the most lagging categories of US services spending on pages 8-15, focused on whether WFH trends and/or activity in central business districts can plausibly explain the gap in spending in each category. The US Services Spending Gap: Key Observations And Conclusions As discussed in greater detail below, we make the following observations about the US services spending gap: Among the seven major categories of US services spending, health care accounts for the largest portion of the services spending gap. Reduced health care spending has little to do with work from home trends, and more to do with an aversion to contracting the disease in a healthcare environment and the reluctance to place elderly relatives in nursing homes given the higher risk that COVID presents to those who are older. Some recreation services spending has been impacted by WFH trends and thus may be permanent, but a lingering fear of crowded indoor spaces and still-recovering international tourism appear to be more important drivers of the recreation services spending gap. Some portion of reduced transportation services spending may be permanent (either in whole or in part), as the spending gap in road transportation seems strongly connected to WFH trends. But the sizeable and impactful decline in real spending on motor vehicle leasing is likely to recover as motor vehicle production improves over the coming year, suggesting that transportation services spending will continue to improve over the coming year relative to its pre-pandemic trend even if a spending gap in this category of services spending is permanent or long-lasting. Personal care and clothing services is mostly responsible for the spending gap in other services, and clear WFH effects do suggest that a reduction in spending in this category may be permanent. However, these categories are relatively small, and in some cases have been partially offset by what is likely to be a permanently positive spending gap on equivalent goods. The takeaway for investors is that the services spending gap currently exists due to a combination of WFH trends and evidence of lasting aversion to COVID-19. While some investors may interpret these observations as suggesting that the gap will act as a permanent or long-lasting drag on consumer spending, we disagree for two important reasons. First, we agree that some form of hybrid work arrangements will be permanent for many businesses, and that a spending gap may be permanent or long-lasting for spending categories most closely tied to WFH effects. But this also suggests that the goods-equivalent spending that has occurred as a result of this decline in services spending will also be permanent. In other words, some of the drag that permanent WFH effects will have on overall consumer spending will be offset by a permanent increase in certain categories of goods spending. Chart II-5Some Of The Permanent Drag On Services Spending Will Be Offset By Permanently Higher Goods Spending Chart II-5 highlights the sum of spending for two pairs of clearly substitutable services/goods categories: miscellaneous personal care services plus personal care products, and sporting equipment, supplies, guns, and ammunition plus membership clubs and participant sports centers. The chart highlights that the sum of these four categories is currently above its pre-pandemic trend, highlighting that permanently lower spending in some services categories affected by WFH trends will likely be offset by permanently higher spending in some goods categories. Second, we doubt that a strong aversion to a COVID-19 infection will be permanent, as the endemicity of the disease has yet to be recognized by the public and normalized by political leaders and health professionals. This is especially true given that the availability and awareness of Pfizer’s Paxlovid antiviral therapy is still in its early stages in the US, and remains severely restricted in other developed economies and (for now) essentially unavailable in the emerging world. As an additional point concerning the lingering societal fear of COVID-19, estimates for the likely annual disease burden from “endemic COVID” are now coming into focus. In a recent New York Times opinion piece, the author cited forecasts from a number of medical professionals that endemic COVID-19 will likely infect roughly half of the US population per year, and will kill on the order of 100,000-250,000 Americans annually.1 That compares with roughly 50,000 fatalities over the course of a year from the worst flu season experienced over the past decade, implying that COVID-19 will end up being between 2-5 times as bad over the longer term as worst-case flu. If the disease burden of endemic COVID-19 ends up being on the higher end of that estimate, then it is likely that an aversion to crowded spaces and shared human settings will be permanent. But we suspect that the eventually-widespread availability of Paxlovid – and other treatment options that have yet to be developed – makes it more likely that annual fatalities will be on the lower end of that range. Chart II-6“Endemic COVID” Will Still Be A Significant Killer, But It Will Not Likely Cause A Permanent Fear Of Crowded Spaces While tragic, a disease with a fatality rate of 30 per 100,000 people (equivalent to 100,000 US deaths per year) will rank behind accidents, chronic lower respiratory diseases (such as bronchitis, emphysema, and asthma), stroke, and just in line with Alzheimer’s disease as a leading cause of death (Chart II-6). It is certainly unwelcome that a new leading cause of death has emerged. But given that COVID-19 will never go away, we doubt that this will be enough to cause a permanent change in public behavior, suggesting that US services spending will return to normal over time. To the extent that some services spending declines are permanent, we expect that to be partially or fully offset by a permanent increase in substitutable goods spending. Investment Conclusions As we discussed in Section 1 of our report, the risk of a US recession is quite elevated. In a non-recessionary scenario, our analysis suggests that the US services spending gap will continue to close, which will provide support for overall consumption as goods spending slows in response to weak real wage growth and higher interest rates. Chart II-7In A Nonrecessionary Scenario, Excess Savings Will Support Services Spending Chart II-7 highlights that the excess savings that have accumulated since the onset of the pandemic – which can be deployed to support spending – have accrued heavily to upper income earners, who are typically responsible for a significant amount of services spending. While it is true that upper income earners have also suffered a significant wealth shock from the combined effect of falling stock and bond prices, we strongly suspect that excess savings and the transition to endemic COVID-19 will support services spending and cause it to move toward the level that would have prevailed had the pandemic not occurred. In a recessionary scenario, we doubt that services spending would fall significantly, given that it is still extraordinarily depressed relative to history. However, some cyclical categories of services spending would decline, and Chart II-1 highlighted that services spending does tend to decline during recessions. The key point for investors is that changes in services spending would not be large enough to cushion a meaningful decline in goods spending were a recession to emerge. While the emergence of a US recession is not yet a foregone conclusion, the risk that it will occur is an important reason supporting our a neutral asset allocation stance. As noted in Section 1 of our report, further signs of an impending recession would cause us to recommend that investors underweight risky assets over the coming 6-12 months. Jonathan LaBerge, CFA Vice President The Bank Credit Analyst Gabriel Di Lullo Research Associate Overall Household Consumption Expenditures for Services Household consumption expenditures for services is composed of seven categories of services spending: Housing and Utilities, Health Care, Transportation Services, Recreation Services, Food Services and Accommodations, Financial Services and Insurance, and Other Services. In order to gauge to what degree services spending is likely to be permanently impaired by the COVID-19 pandemic, we estimate the “services spending gap” for each of these seven categories based on the pre-pandemic trend of overall services spending and the pre-pandemic weight of each category (Chart II-8). Chart II-8The Services Spending Gap Is Fairly Broad-Based Spending fell in all seven services categories during the early phase of the COVID-19 pandemic, but the pace of their respective recoveries has been varied. Spending in many of these sectors has not yet fully recovered relative to its pre-pandemic trend (Charts II-9 and 10), contributing to a spending gap of more than $350 billion real dollars.2 Chart II-8 presents a breakdown of this spending gap by category, and we analyze the drivers of each of these gaps by examining subcategories of services spending on pages 8-15. Our subcategory analysis focuses on areas of services spending that are well below their pre-pandemic level, rather than relative to the hypothetical level of spending that would have prevailed had the pandemic not occurred. This is due to BEA data limitations that prevent us from accurately attributing category spending gaps to subcategories in real terms. Charts II-8-10 underscore that the services spending gap is very broad-based. However, four categories stand out as being particularly impactful: health care, recreation services, transportation services and other services. We discuss the causes of the spending gap in these four categories below, with the goal of determining whether they will likely abate as the pandemic continues to recede, or whether they are likely to be permanent. Chart II-9Four Categories Of Services Spending Stand Out… Chart II-10…As Being Particularly Impactful Drivers Of The Services Spending Gap Health Care Real US personal consumption on health care services is currently $126 billion below our estimate of its pre-pandemic trend, and is currently just below its pre-pandemic level (Chart II-11). “Missing” health care spending accounts for the largest share of the overall spending gap for household consumption expenditures for services. Chart II-11“Missing” Health Care Spending Accounts For A Large Part Of The Overall Services Spending Gap Health care spending initially experienced a V-shaped recovery following the onset of the pandemic, but the pace of recovery has since slowed. The sectors displaying the most significant deviations from their pre-pandemic levels are physician services, dental services, and nursing home spending (Chart II-12). The gap in spending on hospital, physician, and dental services is clearly related to the COVID-19 pandemic, in the sense that some households likely fear contracting the disease in a healthcare setting (especially given the invasive nature of dental treatments). It is also possible that households have been visiting doctor and dentist offices less frequently due to work-from-home policies, in cases where these offices were located in or adjacent to central business districts. Nursing home spending is very much the outlier in the health care sub-sectors, in the sense that its recovery has been more U-shaped than V-shaped. As the pandemic placed the elderly at great risk, we suspect that many family members decided to remove them from nursing homes (or postpone moving them into a nursing home), due to the concern that a communal living environment significantly increased the risk of COVID exposure. Bottom Line: We strongly doubt that the gap in healthcare services spending is permanent. The increasing availability of Paxlovid should help physician services, dental services, and nursing home spending recover, although it is possible that nursing home spending will be the most lagging of the three. Still, we expect that the health care services spending gap will close meaningfully over the coming year if a US recession is avoided (and possibly even if a recession does occur). Chart II-12Some Households Likely Fear Contracting COVID In A Healthcare Setting Chart II-13Lingering Fears Of Crowded Indoor Spaces And Still Weak Tourism Explain Weak Recreation Services Spending Recreation Services Real spending on recreation services is currently $75 billion below its pre-pandemic trend, and remains well below its pre-pandemic level (Chart II-14). Despite only accounting for 6% of household consumption expenditure for services, the sharp decline in spending in certain sub-sectors of recreation services has been large enough to significantly contribute to the overall services spending gap. Chart II-14The Recreation Services Spending Gap: Concerts, Amusement Parks, Movies, And Gyms Chart II-13 highlights that the sectors most responsible for the gap in recreation services spending are 1) live entertainment excluding sports, 2) amusement parks, campgrounds and related recreational services, 3) motion picture theatres, and 4) membership clubs and participant sports centers. A fairly clear narrative explains large spending gaps in three of these categories. In contrast to real spending on spectator sports, which is currently $9 billion above its pre-pandemic level, movies and concerts tend to be held indoors, underscoring that large spending gaps in these categories likely reflect lingering fears of contracting COVID in crowded indoor spaces. Membership clubs and participant sports centers spending is also explained by the COVID-fear effect, although some of the spending gap in this subcategory may be long-lasting as it is also seemingly related to work-from-home effects (for example, substituting home exercise equipment for gym memberships). Real spending on amusement parks, campgrounds and related recreational services is somewhat more difficult to explain, given that spending on these types of services tend to occur outdoors. In addition, some high-profile examples of amusement parks, such as those maintained by the Walt Disney Company in California and Florida, have seemingly experienced strong attendance compared with pre-pandemic levels. We suspect that weakness in this spending category reflects the fact that international tourism has yet to return to its pre-pandemic level. Over the past 12 months, visitor arrivals to the US, while rising, have been less than 40% of what prevailed prior to the pandemic. Bottom Line: We strongly doubt that a sizeable majority of the recreation services spending gap is permanent. As noted for healthcare spending, the increased availability of Paxlovid should progressively reduce the fear associated with crowded indoor spaces, which we believe will cause the recreation services spending gap to close meaningfully over the coming year if a US recession is avoided. Transportation Services Real spending on transportation services is currently $64 billion below our estimate of its pre-pandemic trend, and remains well below its pre-pandemic level (Chart II-15). Chart II-15Road Transportation And Motor Vehicle Leasing Are The Largest Contributors To The Transportation Services Spending Gap Similar to recreation services spending, transportation services spending accounts for only 5% of household consumption expenditure for services, but the extent of the decline in certain categories of transportation services spending has significantly contributed to the overall gap in services spending. The sectors responsible for the transportation services spending gap are: road transportation, motor vehicle leasing, motor vehicle maintenance and repair, and parking fees and tolls (Chart II-16). Some of the gap in transportation services spending is related to work-from-home trends, and as such may be permanent (either in whole or in part). The decline in road transportation spending has been heavily driven by a collapse in spending on intercity buses and mass transit, which is strongly connected to reduced office building occupancy in major US cities and also appears to explain reduced spending on parking fees and tolls. In addition, weak motor vehicle maintenance and repair seems strongly correlated with retail and recreation mobility, which remains below its pre-pandemic level. However, reduced spending on motor vehicle leasing accounts for an important portion of the transportation services spending gap, and does not appear to be caused by work-from-home trends. Instead, the decline in leasing seems strongly linked to the decline in motor vehicle inventory that has caused an enormous rise in new and used car prices. As we have discussed at length in previous reports, this decline in vehicle production and sales has been caused by a semiconductor shortage that will eventually abate, underscoring that this subcomponent of transportation services spending will eventually recover. Bottom Line: We expect the transportation services spending gap to close further over the coming year, even if it does not close fully. Some portion of reduced transportation services spending may be permanent (either in whole or in part), but spending on motor vehicle leasing will not be, suggesting that transportation services spending will continue to improve over the coming year relative to its pre-pandemic trend if a contraction in the US economy is avoided. Chart II-16Some Of The Gap In Transportation Services Spending May Be Permanent Chart II-17Personal Care And Clothing Services Spending Has Definitely Been Impacted By Work-From-Home Trends Other Services Other services spending represents a 14% share of household consumption expenditure for services. Real spending on other services is currently $51 billion below our estimate of its pre-pandemic trend, and still below its pre-pandemic level (Chart II-18). In percentage terms, the other services spending gap is smaller than for health care, recreation services, and transportation services, but it is closer in dollar terms because other services spending is a larger expenditure category. Chart II-18Some Other Services Spending Is Higher Than Before The Pandemic, But Personal Care And Clothing Services Is The Laggard Real spending on other services is below its pre-pandemic level in four subcategories: personal care and clothing services, education services, household maintenance, and social services and religious activities. However, the majority of the spending gap in other services is accounted for by personal care and clothing services (Chart II-17). Some components of personal care and clothing services spending are likely permanently impaired (in whole or in part). Almost all of clothing and footwear services spending is made up by spending on laundry and dry-cleaning services, which remains 12% below its pre-pandemic level and is not exhibiting any meaningful uptrend. In addition, within personal care services, spending on hairdressing salons and personal grooming establishments remains well below its pre-pandemic level, and is only slowly recovering in line with central business district office occupancy. However, one interesting aspect of personal care services spending is that spending on personal care products has increased significantly during the pandemic as spending on miscellaneous personal care services decreased. This suggests that any permanently negative spending gap on personal care and clothing services will be at least partially offset by a permanently positive spending gap on personal care products. Bottom Line: Some of the negative spending gap on other services is likely to be permanent or long-lasting due to persistent work-from-home effects, but at least some of this negative gap will be offset by a permanently positive spending gap on the goods equivalent of these services. Footnotes 1 New York Times Opinion, Endemic Covid-19 Looks Pretty Brutal, July 20, 2022 2 Please note that all real dollar references in this report refer to chained (2012) dollars.