Do Canadian Civil Servants Care About the Health of Populations?

Abstract

Objectives. This article describes Canadian civil servants’ awareness of, attitudes toward, and self-reported use of ideas about the determinants of health.

Methods. Federal and provincial civil servants in departments of finance, labor, social services, and health were surveyed.

Results. With civil servants in finance departments a notable exception, most Canadian civil servants see the health of populations as a relevant outcome for their sectors. Many (65%) report that ideas about the determinants of health have already influenced policymaking in their sector, but most (83%) say they need more information about the health consequences of the policy alternatives their departments face.

Conclusions. Civil servants should consider developing accountability structures for health and researchers should consider producing and transferring more policy-relevant research.

The most recent incarnation of ideas about the determinants of health has been embraced for more than 2 decades by many people in the health sector. Pick almost any country, and people in the health sector of that country can point to one or more prominent documents that have given expression to these ideas. In Canada, it is the Lalonde Report1; in the United States, Healthy People and Healthy People 20002,3; and in the United Kingdom, the “Black Report.”4 These documents highlight the ways in which the environments in which people live, work, and play can have profound implications for the health of populations.

But after all this time, we don’t know whether and how these ideas have taken root outside the health sector, where most of their policy implications lie. For example, tax policies, labor market policies, and early childhood development policies, while developed with explicit economic or social objectives, very likely have profound health consequences.5–7 This leads us to ask to what extent are civil servants in departments of finance, labor, and social services aware of and disposed toward these ideas, and whether they believe these ideas have influenced policies in their respective sectors.

Environmental impact assessments provide a helpful analogy. Civil servants have grown accustomed to considering the environmental consequences of public policies in addition to their explicit economic or social objectives. Are they considering health consequences in the same way? Many policy actors and social movements support a focus on the health of populations, although they are perhaps more preoccupied with vexing issues in their own sectors (such as access to health care) than those who support a focus on the environment.8 And in the case of some health problems, such as infectious diseases, the poor health of a fellow citizen can affect all of us in as direct a way as a polluted environment.

We believe that important lessons can be derived from an assessment of the extent to which Canadian researchers and civil servants in health departments have been able to transfer ideas about the determinants of health to civil servants outside the health sector. The systematic and widespread nature of these efforts makes Canada a helpful test case. Do these civil servants see the health of populations as a relevant outcome for their sectors? Are ideas about the determinants of health sufficiently well developed to have influenced sectors other than the health sector? If not, what is needed to increase the policy relevance and applicability of these ideas? The answers to these questions, which may differ by level of government or by sector, can inform decisions in other countries about whether and how to facilitate the further development and uptake of these ideas.

METHODS

A survey was mailed to 153 civil servants in Canadian federal and provincial departments of finance, labor, social services, and health, from July through December 2000. The term “civil servants” includes all individuals with policy (not program or corporate management) functions at the second-from-top rank within the politically neutral civil service (assistant deputy ministers) and, for the federal government, at the third-from-top rank (directors general) as well. Civil servants in these 4 sectors can, through their influence on the economy, labor markets, social programs, and the health care system, affect many of the determinants of health and, through them, the health of populations.

Survey instrument

The survey instrument consisted of a series of statements assessing civil servants’ awareness of, attitudes toward, and use of ideas about the determinants of health, as well as their sources of knowledge. All statements were derived from published interview and survey instruments, government documents, research papers, and our own interviews with civil servants and researchers.9–11 For example, the categorization of health determinants used in the awareness statements was developed by Canada’s federal government.12 All statements were followed by a 5-point Likert scale, with the scale for awareness statements defined by familiarity; the scale for attitude and use statements, by agreement; and the scale for knowledge sources, by importance. The survey instrument was pilot-tested with civil servants from both levels of government and all 4 sectors and was modified to reduce errors attributable to the way in which statements were worded or the order in which questions appeared. We cannot, however, fully address the concern that respondents may have provided answers that they perceived to be socially desirable.

Analyses

Descriptive analyses were conducted for all civil servants together, then by level of government (federal or provincial) and by sector (finance, labor, social services, or health). For awareness statements, we calculated the proportion of respondents who were quite familiar or very familiar with a particular topic or document. For statements about attitudes toward and use of these ideas, we calculated the proportion of civil respondents who mildly or strongly agreed with a particular statement.

We hypothesized that a higher proportion of federal than provincial civil servants would be aware of and disposed toward ideas about the determinants of health and would report having used these ideas. Unlike their provincial counterparts, federal civil servants are not exposed to the demands of administering a health care system, and they are more likely to have been exposed to documents about the determinants of health produced by their own government. We also hypothesized that civil servants in the health sector would be more familiar with and disposed towards these ideas than civil servants in other sectors. We therefore performed pairwise t tests to assess differences by level of government, with federal as the reference category, and by sector, with health as the reference category. To identify differences across sectors and not just between health and other sectors, we performed F tests to determine whether the proportions for any sector were different from the mean proportion for all sectors.

RESULTS

The overall response rate for the survey was 74% (Table 1 ▶). Response rates were roughly similar across the 2 levels of government (federal and provincial) and across 3 of the 4 policy sectors (labor, social services, and health). The response rate for civil servants from finance departments was low (45%). Survey respondents had similar age, sex, work experience, and training profiles across the 2 levels of government, but different sex, work experience, and training profiles across sectors.

TABLE 1—

Characteristics of Respondents to the McMaster University Survey of Canadian Civil Servants

		Level of Government		Sector
	Overall	Provincial	Federal	Finance	Labor	Social Services	Health	F-test P
Completed survey, no.	113	86	27	13	36	32	32	NA
Received survey, no.	153	114	39	29	43	41	40	NA
Response rate, %	74	75	69	45	84	78	80	NA
Age 30–50 y (vs >50 y), %	60	63	52	77*	66	57	50	.33
Female, %	30	30	30	8	26	50**	25	.02
Government health sector work experience, %	51	50	56	15***	24***	46***	100	.00
Health care delivery work experience, %	32	35	22	0***	9***	32***	70	.00
Economics/finance training, %	67	63	78	92*	74	50	64	.04
Health-related training, %	41	40	44	8***	21***	37***	77	.00
Research training, %	55	54	58	39	69	46	55	.20

Note. NA = not applicable. Pairwise t tests were conducted to assess differences by level of government (with federal as the reference category) and by sector (with health sector as the reference category). F tests of the independence of proportions were conducted to identify differences among sectors.

Source. McMaster University Survey of Canadian Civil Servants, 2000.

*P ≤ .10;

** P ≤ .05;

*** P ≤ .01.

Awareness

More than half the respondents reported that they were quite familiar or very familiar with ideas about the determinants of health (Table 2 ▶). As well, more than half the respondents were quite familiar or very familiar with research related to the impact of specific “determinants of health” (as categorized by Health Canada) on the health of populations, with the exception of the category called biology and genetic endowment. Fewer than one third, and in some cases fewer than one fifth, reported being quite familiar or very familiar with landmark Canadian documents12–16 that have contributed to ideas about the determinants of health. These topics and documents were equally familiar (or unfamiliar) to federal and provincial civil servants, with the exception of the federal government’s Lalonde Report,1 which was better known to federal civil servants. Civil servants from finance departments had consistently lower awareness levels than those from health departments.

TABLE 2—

Canadian Civil Servants’ Awareness of Ideas and Documents About Determinants of Health: Percentage Responding “Quite Familiar” or “Very Familiar” vs “Not At All Familiar,” “Not Very Familiar,” or “Moderately Familiar”

		Level of Government		Sector
	Overall	Provincial	Federal	Finance	Labor	Social Services	Health	F-test P
Ideas about determinants of health	58	54	70	23***	32***	62***	94	.00
Research related to impact of determinants of health (as categorized by Health Canada) on the health of populations
Healthy child development	71	70	73	42***	63*	80	81	.03
Employment/working conditions	67	67	65	17***	77	70	72	.00
Income and social status	66	61*	81	25***		80	75	.00
Education	65	63	73	33***	66	63	78	.05
Health services	59	59	58	50*	43***	57**	81	.01
Physical environment	58	53*	73	25**	63	57	66	.09
Personal health practices and coping skills	57	57	58	25***	57	50**	75	.02
Social support networks	51	47	65	8***	34***	67	72	.00
Biology and genetic endowment	35	33	42	25	37	30	41	.72
Documents that have contributed to ideas about determinants of health in Canada
Lalonde Report1	32	23***	62	0***	6***	40**	66	.00
Strategies for Population Health13	31	26*	46	0***	8***	40*	59	.00
Why Are Some People Healthy and Others Not?14	26	23	35	0***	6***	34	50	.00
Ottawa Charter for Health Promotion15	18	15	27	0***	0***	17***	47	.00
Nurturing Health16	18	16	27	0***	3***	17***	44	.00

Note. Pairwise t tests were conducted to assess differences by level of government (with federal as the reference category) and by sector (with health sector as the reference category). F tests of the independence of proportions were conducted to identify differences among sectors.

Source. McMaster University Survey of Canadian Civil Servants, 2000.

*P ≤ .10;

** P ≤ .05;

*** P ≤ .01.

Attitudes

Only 5% of the respondents felt that knowledge about the determinants of health has little practical policy application, and 85% agreed that health determinants should be considered in all government initiatives (Table 3 ▶). Most (83%) felt they needed more practical information about effective interventions that governments can undertake in order to give greater attention to initiatives that will improve the health of populations. Fewer than half the respondents felt that jurisdictional responsibility for government action to improve the health of populations should rest with government health departments. More respondents supported investments in the prekindergarten years and in data collection and research than supported other possible investments.

TABLE 3—

Canadian Civil Servants’ Attitudes Toward Ideas About Determinants of Health: Percentage Responding “Mildly Agree” or “Strongly Agree” vs “Strongly Disagree,” “Mildly Disagree,” or “Neutral/Undecided”

		Level of Government		Sector
	Overall	Provincial	Federal	Finance	Labor	Social Services	Health	F-test P
Ideas not policy-relevant or -applicable
Knowledge has little practical policy application	5	7	0	8	6	0	9	.42
Determinants shaped by global economic forces	9	7	15	0*	11	3*	16	.22
Decisions must be guided by shorter-term considerations	15	17	11	15	6***	13*	29	.07
More important considerations than impact on health of populations	35	33	42	58**	40	35	22	.13
Nongovernment groups responsible for many possible initiatives	89	87	96	85	80**	94	97	.12
In general terms, here’s what should be done
Consider health determinants in all government initiatives	85	85	85	54***	80	94	94	.00
Provide practical information on effective interventions	83	76**	96	85	71**	87	91	.18
Create new cross-sectoral decisionmaking structures	68	73**	52	31***	57**	83	81	.00
Make health impact assessments mandatory	61	57	70	15***	49***	66**	88	.00
Leave jurisdictional responsibility with health department	45	43	54	58	46	38	47	.69
In specific terms, here’s what should be done
Invest more in the prekindergarten years	83	79**	93	62**	76	97	88	.02
Invest more in research and data collection	80	80	81	46***	83	87	84	.01
Invest more in increasing social cohesion in communities	65	61	77	33**	69	69	69	.12
Invest more in reducing job insecurity and job strain	60	59	62	25***	60	61	72	.05
Improve economic prosperity more than reduce inequalities	30	30	30	62***	21	34	22	.03

Note. Pairwise t tests were conducted to assess differences by level of government (with federal as the reference category) and by sector (with health sector as the reference category). F tests of the independence of proportions were conducted to identify differences among sectors.

Source. McMaster University Survey of Canadian Civil Servants, 2000.

*P ≤ .10; ** P ≤ .05; *** P ≤ .01.

Attitudes toward the determinants of health were generally similar across levels of government, with some notable exceptions. Provincial civil servants were less likely than federal civil servants to say they needed information on effective interventions or to support investments in the prekindergarten years, but they were more likely to support the creation of new cross-sectoral decisionmaking structures. Support for cross-sectoral decisionmaking structures came particularly from civil servants in the social services and health sectors, and support for mandatory health impact assessments came particularly from civil servants in the health sector.

Civil servants in finance departments were outliers in many of their attitudes. They were less convinced than their counterparts in other sectors that health determinants should be considered in all major government initiatives. They were more supportive than their counterparts of improving economic prosperity rather than reducing inequalities. Civil servants in finance departments were less supportive than those in other sectors about investments in any type of policy action.

Use in Policymaking

About two thirds of the respondents felt that knowledge about the determinants of health had influenced the development of policies or programs in their respective sectors (Table 4 ▶). The same proportion felt that they were always trying to improve the health of populations, even though they don’t use that language. A much higher proportion agreed with the positively framed statement about how this knowledge had been used within the health sector than agreed with the negatively framed statement. Federal and provincial civil servants shared similar views about whether and how these ideas have been used. Civil servants from departments of finance and labor tended to be less likely than their counterparts in social services and health to see themselves as always trying to improve the health of populations.

TABLE 4—

Canadian Civil Servants’ Assessments of How Ideas About Determinants of Health Have Been Used: Percentage Responding “Mildly Agree” or “Strongly Agree” vs “Strongly Disagree,” “Mildly Disagree,” or “Neutral/Undecided”

		Level of Government		Sector
	Overall	Provincial	Federal	Finance	Labor	Social Services	Health	F-test P
Knowledge pushed us to consider the value of health care spending practices	69	70	67	69	80*	72	53	.12
Knowledge influenced development of policies in my sector	65	62	74	46	63	81*	59	.12
Always trying to improve the health of populations, just don’t use that language	65	62	73	42*	51*	81	72	.02
Focus used to justify inappropriate health sector funding cuts	15	12	23	8*	0***	21	28	.01

Note. Pairwise t tests were conducted to assess differences by level of government (with federal as the reference category) and by sector (with health sector as the reference category). F tests of the independence of proportions were conducted to identify differences among sectors.

Source. McMaster University Survey of Canadian Civil Servants, 2000.

*P ≤ .10; ** P ≤ .05; *** P ≤ .01.

Sources of Knowledge

Fewer than half the respondents cited research articles or books, professional activities (e.g., conferences, meetings, and briefings), and the media as important or very important sources of knowledge about the determinants of health, whereas more than half cited government or professional reports and informal discussions with colleagues and professional contacts (results not shown). Civil servants from departments of finance and labor were less likely to rely on research articles or books, government or professional reports, and professional activities than civil servants from health departments. Civil servants from departments of labor were particularly likely to cite the media as an important source of knowledge about the determinants of health.

DISCUSSION

Civil servants in labor and social services departments see the health of populations as a relevant outcome for their sectors. We take as evidence of this the fact that 86% of the respondents from these departments agreed with the statement that health determinants should be considered in all government initiatives and only 38% agreed with the statement that there are more important considerations in policymaking than a policy’s impact on the health of populations. As well, 42% of these civil servants were not willing to see their responsibilities for improving the health of populations left with health departments. And yet we have seen just this type of delegation of responsibilities with, for example, Healthy People 2000, which identified priority lead agencies exclusively from within the health sector.3(p659)

Civil servants in finance departments felt quite differently. While we expected civil servants in the federal government and in health sectors to be the outliers, civil servants in finance departments emerged as the outliers. Our follow-up telephone calls to them were often greeted with, “This isn’t relevant to me.” The 45% of civil servants in finance departments who completed the surveys did not believe that they should consider health determinants in all major government initiatives. As well, compared with civil servants in other sectors, they were much less aware of ideas about the determinants of health in general and of research about the impact of specific health determinants on the health of populations. Civil servants in finance departments were also less likely to want to invest more in research and data collection or to undertake policy initiatives that could influence specific health determinants.

Many civil servants outside the health sector reported that knowledge about the determinants of health has already influenced the development of policies in their sector. We cannot assess this influence relative to other factors, such as the values of the governing party, the power of interest groups, and the conditioning effects of particular institutional arrangements.16 That said, the correspondence between the particular expression of this knowledge in Canada and the particular types of influence reported by the civil servants we surveyed lends credibility to these findings. The focus in Canada on efficiency in the production of health8 (as opposed to the focus on the socioeconomic gradient in the United States17 and the focus on health inequalities in the United Kingdom18) appears to be reflected in the high proportion of civil servants agreeing that this knowledge pushed them to consider the value of health care spending practices. And the focus in Canada on early childhood development19–21 appears to be reflected in the high proportion of civil servants in social services departments agreeing that this knowledge has influenced the development of policies and programs in their sector.

Respondents to our survey generally agreed that they need more information about the health consequences of the policy alternatives that their departments face. While support for more research and data collection was lukewarm in finance departments, civil servants in other sectors appeared willing to make the necessary investments. Calling for more research is a familiar “tactical” use of research22; it suggests a commitment to action while removing the obligation for meaningful and immediate action. But at the same time, calling for more research indicates that, after more than 2 decades of exposure to ideas about the determinants of health, civil servants feel they need to signal something.23 Additional research is needed to ascertain whether elected officials feel similarly.

Our suggestions about what these survey results mean for civil servants are ambitious. Civil servants need to think seriously about how they can, through their influence on the economy, labor markets, social programs, and the health care system, affect the determinants of health and, through them, the health of populations. Perhaps there are lessons for these civil servants from the environmental field. Civil servants in the environmental sector have been far more successful at putting in place accountability structures that require their colleagues to make explicit the trade-offs between economic or social objectives and environmental consequences. Civil servants will be seeing more and more health research that has implications outside the health sector. It makes sense to begin thinking now about accountability structures for the health of populations.

Our suggestions about what these survey results mean for research and for knowledge transfer are equally ambitious. First, research funders should begin to support the evaluation of interventions and natural experiments that involve health determinants, at the same time as they continue to support research that explores relationships between possible health determinants and the health of populations. This research should attempt to identify the health consequences of policies as well as the extent to which the policies achieve their economic or social objectives. Second, researchers should begin to distill the key messages arising from research on the determinants of health and take concrete steps to use these messages to educate civil servants, especially in hard-to-reach finance departments. We recognize, however, that Homer Simpson’s adage may apply to civil servants in finance departments: “Just because I don’t care, doesn’t mean I don’t understand.”