A “health in all policies” review of Canadian public finance

Résumé

Objectif

Il est démontré que la santé est principalement le fruit de ses déterminants sociaux, et comme de fait, la recherche sur les systèmes de santé montre que les dépenses publiques relatives aux programmes sociaux sont souvent plus fortement corrélées à la santé des populations que les investissements dans les soins médicaux. Notre étude vise à aider les Cabinets provinciaux et fédéraux du Canada à en prendre acte en introduisant le concept de « la santé dans toutes les politiques » (Health in All Policies, ou HiAP) dans les débats budgétaires.

Méthode

L’étude est descriptive; elle analyse des données secondaires accessibles au public sur les budgets fédéraux et provinciaux pour déterminer comment le financement public des investissements dans les déterminants sociaux de la santé (DSS) aux stades précoces (< 45 ans) et ultérieurs (65 ans et plus) du parcours de vie a évolué depuis 1976 par rapport aux investissements dans les soins médicaux.

Résultats

Les dépenses en soins médicaux ont augmenté de 3 983 $ par personne de 65 ans et plus depuis 1976. Cette augmentation dépasse de 45 % l’augmentation combinée des dépenses en services de garde, en congés parentaux, en aide au revenu familial, en éducation et en soins médicaux par personne pour les moins de 45 ans. De toutes les nouvelles dépenses pour les Canadiens plus jeunes, les soins médicaux ont reçu les investissements les plus importants. Alors que les dépenses médicales pour les retraités ont dépassé d’un peu plus de la moitié le rythme des dépenses en revenus de retraite, les dépenses médicales pour les Canadiens plus jeunes ont augmenté presque autant que les dépenses pour l’ensemble des politiques de DSS à leur endroit.

Conclusion

Depuis 1976, il y a une plus grande concordance entre l’approche HiAP et le financement public du Canada pour les aînés que pour les Canadiens plus jeunes. Ces résultats offrent aux décideurs d’importantes informations rétrospectives pour évaluer les futurs investissements publics dans les soins médicaux et les déterminants sociaux de la santé pour tout le parcours de vie, ainsi que les plans de financement de ces investissements.

Introduction

While physicians perform important work to treat illness, research shows that government spending on social programs often has stronger associations with population health than does public investment in medical care. A higher social/medical spending ratio associates with greater life expectancy, lower infant mortality, and fewer potential years of lost life among OECD countries (Bradley et al. 2011). Similar results are found across US states (Bradley et al. 2016). In Canada, Dutton et al. (2018) show that a 1% increase in social spending per dollar spent on medical care is associated with a 0.1% decrease in potentially avoidable mortality and 0.01% increase in life expectancy. These findings align with extensive research showing that social determinants of health (SDOH) exert stronger influence over population health than does the medical system (Hood et al. 2016; Senate Subcommittee on Population Health 2009; WHO Commission on the Social Determinants of Health 2008).

Such evidence encourages Canadian governments to revisit whether they invest in “health in all policies” (HiAP) in recognition that “health is not created by health service provision alone but largely also by determinants of health that together affect the health of individuals and communities” (Ollila et al., 2013, p. 4). HiAP emphasizes that government budget priority setting has important health implications by determining resource allocation between medical and non-medical ministries, and encourages increased accountability for health impacts throughout government by monitoring interministerial budget trends (Kershaw 2018a). There is growing concern in the literature that medical spending crowds out spending on the social determinants given other budget constraints (Bradley et al. 2016; Watkins et al. 2017; Tran et al. 2017; McCullough et al. 2012). In the United States, Steuerle and Isaacs (2014, p. 2214) warn this risk is particularly great for spending on children as a result of “automatic growth in health, retirement, and tax subsidy programs, along with the failure of revenues to keep pace with the overall growth in spending” amid the aging of the population. Their findings raise questions about whether similar trade-offs occur in Canadian budgets.

Objective

This study aims to support Canadian provincial and federal cabinets to engage in a HiAP analysis during budget deliberations to monitor trends in medical and SDoH spending, paying particular attention to finance trends relevant for the generations raising children and for seniors. The study asks: How has public finance for SDoH investments in earlier and later life course stages evolved over the last four decades (1976–2016) relative to investments in medical care? I explore this question in the context of the demographic bulge associated with Canada’s aging population, and examine whether contemporary spending increases have come from general revenue, or from pools of revenue to which residents contribute specifically in anticipation of increased expenditures that occur later in life. Results provide decision-makers with important retrospective information by which to evaluate future investments in and beyond medical care, across the life course, and the design of revenue plans to finance those investments.

Methods

This study is descriptive, analyzing secondary, publicly available data about federal and provincial budgets. I focus on government spending for retirement income and medical care for older age cohorts, along with spending on childcare services, parental leave, cash supports for families with children, education, and medical care for younger cohorts (see Table 1 for data sources). Literature about the social determinants of health confirms that these represent key policy levers (Mikkonen and Raphael 2010), although it is not an exhaustive list.

Table 1.

Data summary and source

Data	Years	Source
Public finance
Government revenue and spending on Old Age Security, C/QPP, family income	1976, 1998, 2016	Statistics Canada Table 36-10-0477-01
Government spending on medical care	1976, 1998, 2016	CIHI (2018a; 2018b)
Government spending on child care	1976	Government of BC 1977, D.41, prorated for BC’s 11% of national population
	1998, 2016	Friendly et al. (2018, Table 13)
Government spending on parental leave	1976	Canadian Tax Foundation, (1979, Table 7–9)
	1998	Government of Canada (1999, 10)
	2016	Government of Canada (n.d., Chart 2)
Government spending on elementary and secondary education	1976, 1998, 2016	Statistics Canada Table: 37-10-0067-01
Government spending on postsecondary education	1976, 1998, 2016	Statistics Canada Table: 36-10-0484-01
Government debt	1976, 1998, 2016	Statistics Canada Table 36-10-0532-01 and Table 36-10-0580-01
Inflation	1976, 1998, 2016	Statistics Canada Table 18-10-0005-01
Gross domestic product (GDP)	1976, 1998, 2016	Statistics Canada Table: 36-10-0103-01
Demographic information
Population age distribution	1976, 1998, 2016	Statistics Canada Table: 17-10-0005-01
Change in % of students earning postsecondary credentials	1976	Statistics Canada (1978a, 1978b)
	1998	Statistics Canada 2001 Census, Catalogue Number 97F0017XCB2001001
	2016	Statistics Canada 2016 Census, Catalogue Number 98-400-X2016241
Change in % female labour force participation	1976, 1998, 2016	Statistics Canada Table 14-10-0018-01

Following other research, I define older cohorts as age 65+ and younger cohorts as under age 45. The latter age aggregation reduces the need to decouple the value of benefits enjoyed by children apart from their adult caregivers (see Kershaw and Anderson (2016) for discussion of this methodological decision). However, I also report sensitivity analyses that apportion social spending for younger Canadians to more targeted age cohorts.

I quantify government spending on medical care and key social policies at three time periods: 1976, 1998, and 2016. The latter is the most recent year for which the Canadian Institute for Health Information (CIHI) (2018a) provides data about the age distribution of medical spending at the time this study was conducted. 1976 marks the beginning of the 5-year period in which the largest part of the Baby Boom generation (born 1946–1964) came of age as young adults. I thus examine government spending at two pivotal stages for Boomers: when raising children 40 years ago, to which I compare public finance now for Canadians under 45, a cohort that includes many of their children; and now at retirement, to which I compare spending 40 years ago for the cohort of seniors that included many of their parents.

I select 1998 for comparison because it marks the first full year after the federal government adjusted Canada Public Pension (CPP) revenue collection policy. This revenue collection adjustment is important for examining the degree to which Canada shares the risk reported in the USA that retirement income and medical spending for seniors has crowded out SDoH investments at earlier life course stages. Prior to the mid-1990s, Canadian governments operated programs on which Canadians draw primarily when older, like CPP, Old Age Security (OAS) and medical care, as pay-as-you-go systems. These systems see governments collect revenue in each year to correspond (more or less) with the cost of benefits paid in the same year. By the mid-1990s, the federal government recognized a shrinking workers/seniors ratio required changes to CPP revenue collection if it were to remain solvent for future generations, and shifted it toward a prepay system. Rather than contribute to general revenue, individuals now contribute payments over their working lives directly to the CPP system. These payments are closer in amount to average projected future benefits (adjusting for returns on the pre-investment). By examining the change in government spending between 1998 and 2016, I examine implications for public finance that flow from the shift to collect more revenue for retirement security from pools of funding to which residents specifically contribute for use later in life, but not for medical care, on which Canadians also draw disproportionately when we are older.

Data about the age distribution of medical care spending are from CIHI (2018a), which provide per capita figures back to 1998 for cohorts under age 1, age 1–4, 5–9, 10–14, etc. to age 90+. Per capita data are multiplied by the Canadian population to calculate the percentage of spending used by residents age 65+ and under age 45, revealing that 30.3% of public medical care spending went to Canadians under age 45 in 2016, compared to 44.8% for those age 65+. The corresponding figures for 1998 are 37.5% for younger Canadians compared to 41.6% for seniors. These percentages are attributed to the total public budget for medical care reported by CIHI (2018b), which provides annual funding figures back to 1975 (see Table 2).

Table 2.

Estimates of % medical spending to < age 45 and age 65+: 2016, 1998

	2016				1998
Age groups	2016 CIHI per capita medical spending	2016 population	Age cohort estimate based on CIHI per capita data	% < 45 and 65+	1998 CIHI per capita medical spending	1998 population	Age cohort estimate based on CIHI per capita data	% < 45 and 65+
< 1	11,973.18	386,729	4,630,377,215	30.3%	5013.71	344,912	1,729,289,169	37.5%
1–4	1713.83	1,555,293	2,665,502,614		783.60	1,527,329	1,196,811,234
5–9	1425.55	1,985,144	2,829,915,617		614.04	2,060,175	1,265,035,993
10–14	1512.81	1,886,340	2,853,678,573		596.20	2,026,885	1,208,419,017
15–19	1861.83	2,066,404	3,847,291,353		783.33	2,052,913	1,608,100,407
20–24	1891.88	2,467,287	4,667,817,700		944.82	2,013,069	1,901,993,740
25–29	2303.14	2,515,993	5,794,689,140		1128.52	2,105,867	2,376,513,421
30–34	2576.82	2,529,348	6,517,667,868		1157.31	2,403,274	2,781,340,262
35–39	2496.74	2,455,403	6,130,501,380		1071.93	2,679,049	2,871,751,012
40–44	2514.62	2,345,732	5,898,626,562		1084.18	2,510,646	2,721,999,222
45–49	2837.88	2,415,365	6,854,518,886		1239.85	2,191,486	2,717,113,274
50–54	3308.98	2,711,448	8,972,117,603		1482.54	1,872,586	2,776,189,151
55–59	3966.63	2,653,893	10,527,000,067		1826.45	1,435,536	2,621,936,616
60–64	4912.44	2,300,327	11,300,221,631		2323.89	1,211,326	2,814,991,947
65–69	6481.46	1,976,211	12,808,738,409	44.8%	3449.05	1,143,673	3,944,580,342	41.6%
70–74	8348.36	1,438,585	12,009,830,898		4768.66	984,757	4,695,969,559
75–79	11,080.64	1,035,621	11,475,340,627		6517.19	756,829	4,932,397,195
80–84	15,441.71	753,852	11,640,764,849		9209.41	467,412	4,304,589,793
85–89	22,711.78	492,434	11,184,052,779		15,743.07	249,405	3,926,400,566
90+	29,248.31	293,195	8,575,458,729		16,144.32	*included in 85–89
Projected total medical spending			151,184,112,498				52,395,421,919
Actual total medical spending per CIHI			162,733,403,556				59,169,391,359

Sources: Population data from Statistics Canada Table 17-10-0005-01, Per capita medical spending data from CIHI (2018a), Actual total public medical spending data from CIHI (2018b)

Since CIHI does not provide per capita spending data before 1998, I rely on two assumptions used elsewhere in the peer-reviewed literature (Kershaw 2018c) to estimate per capita medical spending values in 1976. The first calculates the average annual change between 1998 and 2016 for each CIHI age group, and attributes that average change to each year before 1998 to estimate 1976 values. Multiplication of these 1976 per capita estimates by the population that year results in 44% of medical spending going to Canadians under age 45, and 36.1% for Canadians age 65+. As a sensitivity analysis, I attribute the per capita spending values in 1998 to the population distribution in 1976. This sensitivity analysis predicts 46.9% of spending in 1976 went to those under age 45, and 32.1% went to those age 65+ (see Table 3).

Table 3.

Estimates of % medical spending to < age 45 and age 65+: 1976

	1976								1976 sensitivity analysis
Age groups	2016 CIHI per capita medical spending	1998 CIHI per capita medical spending	Total per capita change 1998–2016 as % of 2016 per capita $	Average annual change = total per capita change divided by 18 years (2016–1998)	1976 per capita estimate = average annual % change attributed to years 1998–1976, inclusive	1976 population	Age cohort estimate based on CIHI per capita data	% < 45 and 65+	1998 CIHI per capita medical spending	1976 population	Age cohort estimate based on CIHI per capita data	% < 45 and 65+
< 1	11,973.18	5013.71	− 58%	− 3.2%	2435	352,901	859,385,598	44.0%	5013.71	352,901	1,769,343,708	46.9%
1–4	1713.83	783.60	− 54%	− 3.0%	400	1,406,416	561,906,635		783.60	1,406,416	1,102,064,106
5–9	1425.55	614.04	− 57%	− 3.2%	303	1,909,617	578,228,132		614.04	1,909,617	1,172,586,910
10–14	1512.81	596.20	− 61%	− 3.4%	281	2,291,715	643,273,383		596.20	2,291,715	1,366,309,380
15–19	1861.83	783.33	− 58%	− 3.2%	381	2,389,251	911,315,173		783.33	2,389,251	1,871,562,753
20–24	1891.88	944.82	− 50%	− 2.8%	508	2,252,144	1,144,104,683		944.82	2,252,144	2,127,877,281
25–29	2303.14	1128.52	− 51%	− 2.8%	600	2,062,690	1,236,856,718		1128.52	2,062,690	2,327,787,304
30–34	2576.82	1157.31	− 55%	− 3.1%	584	1,682,190	982,542,020		1157.31	1,682,190	1,946,820,369
35–39	2496.74	1071.93	− 57%	− 3.2%	528	1,354,239	714,566,216		1071.93	1,354,239	1,451,648,409
40–44	2514.62	1084.18	− 57%	− 3.2%	535	1,286,062	687,928,133		1084.18	1,286,062	1,394,326,306
45–49	2837.88	1239.85	− 56%	− 3.1%	616	1,266,848	780,582,048		1239.85	1,266,848	1,570,701,121
50–54	3308.98	1482.54	− 55%	− 3.1%	747	1,229,164	918,430,846		1482.54	1,229,164	1,822,288,408
55–59	3966.63	1826.45	− 54%	− 3.0%	935	1,030,853	963,899,592		1826.45	1,030,853	1,882,802,818
60–64	4912.44	2323.89	− 53%	− 2.9%	1209	913,129	1,103,753,095		2323.89	913,129	2,122,014,042
65–69	6481.46	3449.05	− 47%	− 2.6%	1932	728,512	1,407,696,359	36.1%	3449.05	728,512	2,512,671,116	32.1%
70–74	8348.36	4768.66	− 43%	− 2.4%	2806	539,115	1,512,580,607		4768.66	539,115	2,570,855,174
75–79	11,080.64	6517.19	− 41%	− 2.3%	3917	366,193	1,434,247,871		6517.19	366,193	2,386,548,779
80–84	15,441.71	9209.41	− 40%	− 2.2%	5592	222,611	1,244,820,507		9209.41	222,611	2,050,116,467
85–89	22,711.78	15,743.07	− 31%	− 1.7%	10,785	113,406	1,223,075,856		15,743.07	113,406	1,785,358,684
90+	29,248.31	16,144.32	− 45%	− 2.5%	9272	*included in 85–89			16,144.32	*included in 85–89
Projected total medical spending							18,909,193,470				35,233,683,138
Actual total medical spending per CIHI							10,817,193,960				10,817,193,960

Sources: Population data from Statistics Canada Table 17-10-0005-01, Per capita medical spending data from CIHI (2018a), Actual total public medical spending data from CIHI (2018b)

Annual data about government revenue and spending on some SDoH policies are available from Statistics Canada sources. However, Statistics Canada does not provide annual spending data for childcare or parental leave. Childcare data in 1976 are from the Government of British Columbia (1977) adjusted for BC’s share of the national population—an assumption used elsewhere in the literature (Kershaw 2018c), because there is limited variation in public finance for childcare among provinces before 1997 when Quebec launched its universal childcare system. Childcare data for 1998 and 2016 are from the Childcare Resource and Research Unit, while parental leave data are collected from documents published by the Canadian Tax Foundation and Government of Canada (see Table 2).

There is no counterpart to CIHI that annually reports the age distribution of social spending. As a result, age analyses of SDoH expenditures entail blunt assumptions (see Table 4). Guided by Kershaw and Anderson (2016) who provide a comprehensive age analysis of contemporary Canadian expenditures, I allocate all spending on Old Age Security (OAS) and the Canada and Quebec Public Pensions (C/QPP) to Canadians age 65+, while conceding some Canadians use C/QPP before they reach age 65. Conversely, I attribute all childcare, parental leave, grade school, postsecondary, and income support for families with children to those under age 45, while conceding that some parents benefitting from these programs are over 45 years, as are some students in postsecondary. For sensitivity analyses, I apportion social spending for younger Canadians to more targeted age cohorts. Given the bluntness of these age assumptions, this study searches for orders-of-magnitude in variation between medical and SDoH spending.

Table 4.

Assumptions about the age distribution of government spending

	Aggregate spending, expressed as share of gross domestic product (GDP) to facilitate comparisons across years			Per capita spending, calculated as aggregate spending/population, reported after adjusting for inflation, economic growth and factors influencing demand
Year	1976	1998	2016	1976	1998	2016
GDP ($ million)	205,123	937,295	2,027,544	Adjusted for 54% increase in GDP/capita in 2016 by comparison with 1976	Adjusted for 20% increase in GDP/capita in 2016 by comparison with 1998
Consumer Price Index				Adjusted by 128.4 (CPI value in 2016)/31.1 (CPI value in 1976)	Adjusted by 128.4 (CPI value in 2016)/91.3 (CPI value in 1998)
Medical care	44.0% to < age 45 (16.99 million people); 36.1% to age 65+ (1.97 million people)	37.5% to < age 45 (19.72 million people); 41.6% to age 65+ (3.60 million people)	30.3% to < age 45 (20.22 million people); 44.8% to age 65+ (5.99 million people)	CIHI per capita data imply a 111% increase in demand among < age 45 and a 49% increase in demand among age 65+ in 2016 by comparison with 1976	CIHI per capita data imply a 54% increase in demand among < age 45 and a 27% increase in demand among age 65+ in 2016 by comparison with 1998
Sensitivity analysis	46.9% to < age 45 (16.99 million people);
	32.1% to age 65+ (1.97 million people)
OAS	100% to age 65+ (1.97 million people)	100% to age 65+ (3.60 million people)	100% to age 65+ (5.99 million people)	Consistent universal use among eligible cohort
CPP	100% to age 65+ (1.97 million people)	100% to age 65+ (3.60 million people)	100% to age 65+ (5.99 million people)	Consistent universal use among eligible cohort
Childcare services	100% to < age 45 (16.99 million people)	100% to < age 45 (19.72 million people)	100% to < age 45 (20.22 million people)	Adjusted for 52% increase in labour force participation for women age 25–44 in 2016 by comparison with 1976	Adjusted for 4% increase in labour force participation for women age 25–44 in 2016 by comparison with 1998
Sensitivity analysis	100% to < age 12 (5.04 million people)	100% to < age 12 (5.15 million people)	100% to < age 12 (5.06 million people)
Parental leave	100% to < age 45 (16.99 million people)	100% to < age 45 (19.72 million people)	100% to < age 45 (20.22 million people)	Adjusted for 52% increase in labour force participation for women age 25–44 in 2016 by comparison with 1976	Adjusted for 4% increase in labour force participation for women age 25–44 in 2016 by comparison with 1998
Sensitivity analysis	100% to < age 1 + primary caregiver (0.71 million people)	100% to < age 1 + primary caregiver (0.69 million people)	100% to < age 1 + primary caregiver (0.77 million people)
Family income support	100% to < age 45 (16.99 million people)	100% to < age 45 (19.72 million people)	100% to < age 45 (20.22 million people)	Consistent universal use among eligible cohort
Elementary and secondary	100% to < age 45 (16.99 million people)	100% to < age 45 (19.72 million people)	100% to < age 45 (20.22 million people)	Consistent universal use among eligible cohort
Sensitivity analysis	100% to age 5–17 (5.63 million people)	100% to age 5–17 (5.32 million people)	100% to age 5–17 (5.11 million people)
Postsecondary	100% to < age 45 (16.99 million people)	100% to < age 45 (19.72 million people)	100% to < age 45 (20.22 million people)	Adjusted for 149% increase in share of age 25–44 who earn postsecondary credentials in 2016 by comparison with 1976	Adjusted for 22% increase in share of age 25–44 who earn postsecondary credentials in 2016 by comparison with 1998
Sensitivity analysis	100% to age 18–44 (9.59 million people)	100% to age 18–44 (12.53 million people)	100% to age 18–44 (13.14 million people)

Sources:

Population data from Statistics Canada Table 17-10-0005-01

Revenue, OAS, C/QPP, Family income from Statistics Canada Table 36-10-0477-01

GDP data from Statistics Canada Table: 36-10-0103-01

Medical care data from CIHI (2018a; 2018b)

1976 childcare data from Government of BC 1977, D.41; 1998 and 2016 data from Friendly et al. (2018, Table 13)

1976 parental leave data from Canadian Tax Foundation (1979, Table 7–9). 1998 data from Government of Canada (1999, 10). 2016 data from Government of Canada (n.d., Chart 2)

Elementary and secondary data from Statistics Canada Table: 37-10-0067-01

Postsecondary spending data from Statistics Canada Table 36-10-0484-01

1976 postsecondary utilization data from Statistics Canada (1978a, 1978b). 1998 data based on Statistics Canada 2001 Census, Catalogue Number 97F0017XCB2001001

2016 postsecondary data from Statistics Canada 2016 Census, Catalogue Number 98–400-X2016241

Female labour force data from Statistics Canada Table 14-10-0018-01

Consumer Price Index data from Statistics Canada Table 18-10-0005-01

Statistical analysis

For the years 1976, 1998, and 2016, I calculate aggregate and per capita spending values for each policy of interest. Aggregate spending changes between these years are calculated as a percentage of gross domestic product (GDP) in order to facilitate comparison across time, because GDP integrates trends related to inflation and economic growth. I explore aggregate spending change values between medical care and SDoH policies within, and across, the two age groups of interest. Changes in spending are analyzed relative to the two categories of revenue: general revenue, and the C/QPP prepay system. The analysis reports implications for public debt.

Per capita expenditure changes are adjusted for inflation and economic growth. The per capita analyses quantify how medical spending has grown relative to key SDoH policies for Canadians over age 65 and under 45 in light of population changes, and changes in demand for the public investments (see Table 4). Per capita investments in childcare and parental leave account for rising demand as a result of the 52% increase in labour force participation among women age 25–44 between 1976 and 2016, and 4% increase from 1998 to 2016. Similarly, postsecondary figures account for the 149% increase in the share of Canadians age 25–44 who earned postsecondary credentials by comparison with 1976, and 22% increase from around 1998 to 2016.1 Changes in demand for publicly funded medical care are integrated in CIHI per capita figures. They imply a 49% increase in use of publicly paid medical care per senior between 1976 and 2016 and a 111% increase per person under age 45. For the period 1998 to 2016, the implied demand increases are 27% for age 65+ and 54% for Canadians under age 45.2

Results

As the number of seniors grew from 2 million to 6 million over the last 4 decades, Table 5 shows that spending on medical care and retirement income increased in aggregate by $87.7 billion. This additional spending is $39.9 billion more than total government revenue increased over the same period. As the number of Canadians under age 45 grew from 17 million to 20.2 million, aggregate spending on the suite of policies for younger Canadians dropped by $15.8 billion. The reduction occurred primarily before 1998, with partial reinvestment thereafter. Government debt levels doubled as a share of the economy, and tripled when measured per capita relative to the population under 45.

Table 5.

Change in aggregate government spending (2016$) on age 65+ and < age 45: 1976, 1998, and 2016

	1976	1998	2016	2016 minus 1976		2016 minus 1998
				%GDP	2016$ millions	%GDP	2016$ millions
GDP ($ millions)	205,123	937,295	2,027,544
Revenue
Total government general revenue	34.99%	43.70%	35.55%	0.56%	11,349	− 8.1%	− 165,184
CPP/QPP revenue	1.60%	2.45%	3.39%	1.80%	36,483	0.9%	19,219
Total	36.59%	46.15%	38.95%	2.36%	47,832	− 7.2%	− 145,965
Spending 65+
From general revenue
Medical care to 65+	1.90%	2.63%	3.59%	1.69%	34,288	1.0%	19,602
Sensitivity analysis	1.69%			1.90%	38,557
OAS	2.10%	2.39%	2.34%	0.24%	4947	0.0%	− 951
General revenue subtotal	4.00%	5.02%	5.94%	1.94%	39,235	0.9%	18,651
From C/QPP revenue	0.54%	1.19%	2.93%	2.39%	48,501	1.7%	35,259
Total (excluding sensitivity analyses)	4.54%	6.21%	8.86%	4.33%	87,736	2.7%	53,910
Spending < 45
Childcare services	0.05%	0.11%	0.23%	0.18%	3587	0.1%	2328
Parental leave	0.07%	0.13%	0.19%	0.12%	2428	0.1%	1277
Family income support	0.95%	0.60%	1.04%	0.09%	1790	0.4%	8747
Elementary and secondary	4.72%	3.82%	3.30%	− 1.41%	− 28,643	− 0.5%	− 10,493
Postsecondary	2.20%	1.86%	2.33%	0.13%	2721	0.5%	9625
Medical care < 45	2.32%	2.37%	2.43%	0.11%	2291	0.1%	1308
Sensitivity analysis	2.47%			− 0.04%	− 827
Total (excluding sensitivity analyses)	10.30%	8.89%	9.52%	− 0.78%	− 15,825	0.6%	12,791
Debt < 45	19.20%	81.77%	43.88%	24.68%	500,405	− 37.9%	− 768,180

Sources:

Population data from Statistics Canada Table 17-10-0005-01

Revenue, OAS, C/QPP, Family income from Statistics Canada Table 36-10-0477-01

GDP data from Statistics Canada Table: 36-10-0103-01

Medical care data from CIHI (2018a; 2018b)

1976 childcare data from Government of BC 1977, D.41; 1998 and 2016 data from Friendly et al. (2018, Table 13)

1976 parental leave data from Canadian Tax Foundation (1979, Table 7–9). 1998 data from Government of Canada (1999, 10). 2016 parental leave data from Government of Canada (n.d., Chart 2)

Elementary and secondary data from Statistics Canada Table: 37-10-0067-01

Postsecondary spending data from Statistics Canada Table 36-10-0484-01

1976 postsecondary utilization data from Statistics Canada (1978a, 1978b). 1998 data based on Statistics Canada 2001 Census, Catalogue Number 97F0017XCB2001001. 2016 data from Statistics Canada 2016 Census, Catalogue Number 98–400-X2016241

Female labour force data from Statistics Canada Table 14-10-0018-01

Debt data from Statistics Canada Table 36-10-0532-01 and Table 36-10-0580-01

Inflation adjustment data from Statistics Canada Table 18-10-0005-01

There are noteworthy differences in public finance between medical care for retirees and retirement income following 1998. Medical spending increased $19.6 billion from a pool of general revenue that dropped by $165 billion. C/QPP spending increased $35.2 billion from a pool of C/QPP revenue that increased by $19.2 billion, and which was subsequently invested to close the gap between revenue and expenditure.

Aggregate spending and demographic changes resulted in per capita medical care investments rising 49% (after inflation) from $8180 per person age 65+ in 1976 to $12,163 in 2016—an increase of $3983 (see Table 6).3 Retirement income spending (OAS plus C/QPP) increased 57% from $11,326 per retiree in 1976 to $17,839 in 2016—an increase of $6513.

Table 6.

Change in per capita government spending (2016$) on age 65+ and < age 45: 1976, 1998, and 2016

	1976		1998		2016	2016 minus 1976		2016 minus 1998
	Inflation adjusted (128.4/31.1)	* 2016 Growth = *1.54	Inflation adjusted (128.4/91.3)	* 2016 Growth = *1.20		w/o Growth	* Growth	w/o Growth	* Growth
GDP ($ per capita)	36,114		46,461		55,876
Revenue
Total government general revenue	12,666	19,553	19,102	22,973	19,866	7200	313	763	− 3108
CPP/QPP revenue	577	891	1069	1286	1897	1319	1005	827	611
Total	13,243	20,444	20,172	24,260	21,762	8519	1318	1590	− 2497
Spending 65+
From general revenue
Medical care to 65+	8180	12,628	9613	11,562	12,163	3983	− 464	2550	602
Sensitivity analysis	7275	11,230				4889	933
OAS	9023	13,929	8745	10,517	7929	− 1094	− 6000	− 816	− 2588
General revenue subtotal	17,203	26,557	18,358	22,078	20,093	2890	− 6464	1734	− 1986
From C/QPP revenue	2303	3556	4351	5232	9910	7606	6354	5559	4677
Total (excluding sensitivity analyses)	19,506	30,112	22,709	27,311	30,002	10,496	− 110	7293	2692
Spending < 45
Child care services	38	58	78	94	226	188	168	148	132
Sensitivity analysis	140	216	323	389	972	832	756	648	583
Parental leave	53	82	88	106	190	137	108	102	84
Sensitivity analysis	1274	1967	3681	4426	4887	3613	2921	1207	461
Family income support	472	729	403	485	1038	566	309	635	553
Elementary and secondary	2352	3630	2554	3072	3314	962	− 316	760	242
Sensitivity analysis	7089	10,944	9472	11,391	13,109	6020	2165	3637	1718
Postsecondary	2725	4207	1508	1813	2338	− 387	− 1869	831	525
Sensitivity analysis	4826	7450	2373	2854	3596	− 1230	− 3854	1223	743
Medical care < 45	1157	1786	1583	1904	2441	1284	655	857	537
Sensitivity analysis	1233	1904				1207	536
Total (excluding sensitivity analyses)	6796	10,493	6214	7473	9547	2750	− 947	3332	2073
Debt < 45	9574	14,779	54,648	65,721	44,013	34,439	29,234	− 10,635	− 21,709

Sources:

Population data from Statistics Canada Table 17-10-0005-01

Revenue, OAS, C/QPP, Family income from Statistics Canada Table 36-10-0477-01

GDP data from Statistics Canada Table: 36-10-0103-01

Medical care data from CIHI (2018a; 2018b)

1976 childcare data from Government of BC 1977, D.41; 1998 and 2016 data from Friendly et al. (2018, Table 13)

1976 parental leave data from Canadian Tax Foundation (1979, Table 7–9). 1998 data from Government of Canada (1999, 10). 2016 data from Government of Canada (n.d., Chart 2)

Elementary and secondary data from Statistics Canada Table: 37-10-0067-01

Postsecondary spending data from Statistics Canada Table 36-10-0484-01

1976 postsecondary utilization data from Statistics Canada (1978a, 1978b). 1998 data based on Statistics Canada 2001 Census, Catalogue Number 97F0017XCB2001001. 2016 data from Statistics Canada 2016 Census, Catalogue Number 98-400-X2016241

Female labour force data from Statistics Canada Table 14-10-0018-01

Debt data from Statistics Canada Table 36-10-0532-01 and Table 36-10-0580-01

Inflation adjustment data from Statistics Canada Table 18-10-0005-01

Three quarters of the per capita increase in retirement spending occurred after 1998, drawing from the growing revenue stream that was created so people prepay in order to reduce the risk that the demographic bulge associated with the aging population would disproportionately absorb general revenue as the Boomers retired. Two thirds of the per capita increase for medical care occurred after 1998, drawing from general revenue, which does not have a prepay component to account for variation in the ratio between retirees and workers. General revenue fell as a share of the economy over the time period. Combined, medical and retirement spending on seniors grew 54% since 1976, on par with economic growth.

Per capita spending on Canadians under age 45 increased 40%, slower than the rate of economic growth (54%), rising from $6797 in 1976 to $9547 in 2016. The increase occurred after 1998. Among the suite of policies, medical investments grew by 111%, or $1284.4 Medical spending increased one third more than grade school funding, more than double family income supports, seven times more than childcare, and nine times more than parental leave, while postsecondary spending declined when measured on a per capita basis. After 1998, medical care remains the largest dollar-value increase in the suite of policies for younger people, but represents only one quarter of the total growth in per capita investment.

I perform several sensitivity analyses reported in Table 6, beginning by apportioning childcare spending entirely to those under age 12 to find a per capita increase of $832 since 1976. When postsecondary spending is allocated only to those age 18–45, there is a per capita reduction of $1230. If parental leave spending is assigned just to children under age one and a primary caregiver, the per capita increase is $3613. If grade school spending is assumed to benefit only children age 5–17, not parental labour force attachment, the per capita increase is $6020. The latter two changes resemble the magnitude of the $3983 increase in medical care per senior, or $6513 combined increase to C/QPP and OAS.

Discussion

The medical spending increases since 1976 are notable for multiple reasons when considered from a HiAP perspective. The first is relative size. The $3983 increase in medical spending per senior is 45% larger than the entire increase per person under age 45 for the suite of policies considered in this study. Medical care also received the largest spending increase in the package for younger Canadians. By comparison, SDoH spending was prioritized more for retirees than for younger Canadians in absolute dollars, and relative to medical investments. Retirement income spending increased $6513—four times more than the $1466 cumulative increase for childcare, parental leave, family income, grade school, and postsecondary. Whereas medical spending for retirees increased just over half the pace of retirement income spending, medical spending for residents under age 45 increased nearly as much as their entire package of SDoH policies considered in this study. These findings imply there has been greater alignment between the HiAP concept and Canadian public finance for seniors than for younger Canadians. This greater alignment coincided with reductions to the prevalence of low income measured by low-income cut-offs (LICO) among seniors from 29% in 1976 to approximately 4.7% in 2016. Children now have similar or higher levels of low income compared with seniors according to all Statistics Canada (Table 11-10-0135-01) measures of this construct.

It is also noteworthy that additional medical spending since 1976 comes from general revenue, whereas increases to retirement income come from C/QPP. Since Canadians began prepaying the latter, larger benefits now enjoyed by seniors partly reflect their larger contributions than past generations. This is not the case for larger medical expenditures, which taxpayers fund in response to annual demand from general revenue. The $34.3 billion aggregate increase in medical spending for Canadians age 65+ is three times larger than the total increase in general revenue since 1976, and medical spending has been growing for young and old alike since 1998 while general revenue has dropped substantially as a share of the economy.

These changes have major implications for general revenue collection, since the ratio of working-age Canadians relative to seniors has dropped from around 7-to-1 in 1976 to less than 4-to-1 today (Statistics Canada 2014). Research published in the Canadian Tax Journal shows that younger Canadians currently pay 42% to 54% more in income taxes to medical care for retirees than today’s retirees paid as young people four decades earlier toward their own elderly. Higher income tax allocations to medical care for contemporary retirees comes despite a reduction in total income tax rates of 12% to 21% over the period (Kershaw 2018b). The two trends erode government fiscal capacity to address risks to the SDoH, especially for contemporary younger cohorts. This helps to explain why, for example, undergraduate tuition has increased 173% since 1976 after inflation (Statistics Canada TLAC Standard Table 8E.1a and Table 37-10-0045-01), while Friendly et al. (2018, Table 7) show childcare fees are generally higher than university tuition.

Given these findings, Canadian decision-makers have reason to revisit the pace of medical spending increases by comparison with other social spending, as encouraged by previous international scholarship. Bradley et al.’s (2016, p. 760) observation that social services in the USA risk being “crowded out to some degree by rising health care costs” applies to Canadian public finance for younger residents more so than for retirees, because per capita investments in later life course stages increased 3.8 times ($10,496/$2750) more since 1976 than for earlier stages. Had spending on programs for Canadians under age 45 kept pace with economic growth as it did for retirees, the total package would have been $947 more per person under age 45 in 2016 or $19.1 billion more when multiplied by all in the age group. As a point of comparison, a high-quality, universal childcare program in Canada would require another $10 billion (Kershaw and Anderson 2009).

Limitations

This study does not consider the age distributional implications that flow from spending on infrastructure, environmental protection and debt financing, nor how such expenditures shape the SDoH. As discussed above, there are also shortcomings in the data sources about social and medical spending, which require blunt assumptions. These include prorating BC provincial spending on childcare to estimate pan-Canadian spending in 1976, along with a number of assumptions about the age distribution of social and medical investments. Given the imprecision, readers should focus on orders-of-magnitude in variation between medical and SDoH spending—not small differences. Further research would be valuable to link age distribution data more systematically to government social spending, as CIHI provides for medical spending.

Conclusion

Just as Steuerle and Isaacs (2014, p. 2214) find that US medical spending is a “mixed blessing” because it has become “the largest” support for generations raising children, while “eat[ing] away at the share of the budgetary pie left for anything else,” the same problem is evident in Canada. The remedy requires recalibration of Canadian public finance to promote health in all policies when budgeting for younger residents: see Dialogue Box A for two examples of how this study’s retrospective findings can be applied to future public finance decisions related to pharmacare and physician remuneration.

Dialogue Box A Implications for recalibrating public finance: two examples

Pharmacare: Public finance insights about the adaptation of revenue sources for C/QPP, but not medical care, should be considered closely as governments advance plans for national pharmacare. Prescription drug costs rose from 2% of the total government medical budget in 1976 to over 8% in 2016 (author calculations based on CIHI 2018b Table A.3.3.1), representing a proportional increase of $10.2 billion in that year ((8.3–2.0%)*$162.7 billion). There are good reasons to build on this investment to create a national pharmacare system that achieves savings from administrative efficiencies and bulk-buying, while also promoting more equitable access (see Wolfson and Morgan 2018). However, the additional $9.7 billion estimated cost could reduce the social/medical spending ratio in Canada, which would be in tension with HiAP evidence suggesting that new investments in social (not medical) policy would achieve greater population health benefits. Any revenue plan for national pharmacare should be informed by this study’s finding that general revenue has decreased since 1998 despite substantial increases in medical spending over that period, especially for the aging population. Consistent with considerations that led the CPP to rely on a prepay revenue collection system in 1998, a pharmacare revenue plan should address intergenerational implications that accompany the launch of a program on which Canadians will draw disproportionately in our later years. This is especially important if pharmacare is implemented as Baby Boomers retire without having contributed to a prepay revenue-collection system while in the labour market. Physician remuneration: As decision-makers consider how to grow the social/medical spending ratio, spending restraint may be more palatable in parts of the medical system that do not directly relate to illness treatment via prescription drugs, diagnostics, hospital stays, etc. Physician remuneration increased from 18.9% of total public medical spending in 1976 to 21.7% in 2016 (author calculations based on CIHI 2018b Table A.3.3.1). This additional portion of the public medical budget cost $4.6 billion in 2016 ((21.7–18.9%)* $162.7 billion). Physician remuneration increases may be in tension with evidence about health promotion if they come at the expense of SDoH investments, and/or when accompanied by Canadian Medical Association lobbying for income tax loopholes that benefit physicians and other high-earners (Vogel 2017) at the expense of general revenue.

The required recalibration of Canadian public finance should be informed by further research about three interrelated themes: (i) normative scholarship about intergenerational justice, as well as applied HiAP scholarship about opportunities to (ii) reallocate existing spending and/or (iii) raise additional revenue. Retrospectively, it will be important to assess whether intergenerational adaptations in public finance over the past several decades have been made in proportion to the SDoH faced contemporarily by different age cohorts, as well as relative to the advantages and disadvantages inherited by those cohorts (e.g., Kershaw 2018c). Prospectively, it will be valuable for researchers and legislators to consider how current public finance trends will influence the projected social/medical spending ratio that future generations will experience across their life course by comparison with cohorts alive today.

As new research supports legislators to examine how intergenerational justice considerations interact with insights about HiAP, decision-makers may be motivated to search for fiscal capacity to adapt policy. While reallocations can be challenging (Smith et al., 2016), governments should consider evidence that “public health and the economy could be served by reallocating medical expenditures to social programs,” as Tran et al. (2017, p. 185) recommend for the USA, and Dutton et al. (2018) imply for Canada. Reallocation options should be examined in concert with proposals for new revenue strategies that respond appropriately to the relative advantages and disadvantages in the SDoH inherited by different age cohorts (e.g., Kershaw 2018b). Applied HiAP scholarship about reallocation and revenue options will be essential for illuminating that there need not be a zero-sum game between younger and older Canadians. New investments for younger generations can be funded from existing spending or added revenue in ways that do not alter public funds available for the aging population.