Written evidence submitted by Dr van Heerde
and Dr Hudson
A. SUMMARY POINTS
1. Findings from recent surveys of public
opinion and development aid show an overwhelming level of support
for the principle of aid to poor countries and official development
aid. These findings have buoyed policymakers who argued that high-levels
of support are necessary for increasing aid levels, notably to
the 0.7% (ODA/GNI ratio) goal for all OECD Development Assistance
Committee (DAC) countries. However, to date, there is no empirical
evidence showing positive covariation between public opinion and
development aid.
2. Our research has shown that using a more
restrictive (and hence valid) measure of support for development
assistance, there is a weak, negative correlation between public
opinion and ODA. Using a more generous measure of support for
aid/concern for poverty/helping the poor, we find a positive,
but very weak correlation. Neither of these results achieve traditional
levels of statistical significance which may be a function of
low sample size, but the conclusion we draw is that based on the
nearly flat slope lines (see figure 1/2 for evidence) there
is no relationship between support and ODA.
3. There dangers to DFID's mission of poverty
alleviation in rendering future policy decisions on development
aid contingent upon results from existing surveys of public support
for development aid programmes. Whilst democratic responsiveness
is key to effective democratic governance, current measures of
public support are subject to extreme measurement error and are
of poor validity.
4. Existing surveys of public attitudes
suffer from poor validity, in other words, they fail to measure
the underlying attitudes they claim to. Moreover, respondents
may understand development aid as essentially humanitarian aid
thus further biasing results. Surveys also fail to report absolute
v. relative support for aid.
5. Measures of support for development aid
do not control for respondent knowledge-levels and perceptions
of aid effectiveness, which results in an upward biasing of levels
of true support.
6. There is little understanding of the
factors that motivate individual-level support for development
aid or concern for poverty which is important in linking domestic
issue publics with development publics, and for developing strategic
communication and the citizenship curriculum.
7. We do not dismiss a (potential) relationship
between public opinion and policy, we suggest here that the public
for development aid is substantially smaller than survey results
would suggest. Future survey instruments need to take into account
existing weaknesses so that we are better able to identify and
understand support for development aid.
B. INTRODUCTION
1. Dr Jennifer van Heerde is Lecturer in
Research Methods in the Department of Political Science/School
of Public Policy, University College London. She received a BA
from the University of California Los Angeles (UCLA) and a Masters
and PhD from the University of California Riverside. Her research
interests include public opinion, political knowledge and communication
and party/campaign finance. Her work has appeared in the British
Journal of Political Science and Journal of Elections Public
Opinion & Parties. A recent paper, co-authored with David
Hudson, on understanding the determinants of individual-level
concern for poverty is forthcoming in Political Studies.
2. Dr David Hudson is Lecturer in International
Relations and Political Economy in the Department of Political
Science/School of Public Policy, University College London. He
received a B.Soc.Sci, MA, and PhD from the University of Birmingham,
UK. His research interests include financing for development,
global financial governance and support for poverty alleviation.
His work has appeared in Review of International Studies,
and Millennium. A book on Global Finance and Development
is forthcoming with Routledge Press. A recent paper, co-authored
with Jennifer van Heerde, on understanding the determinants of
individual-level concern for poverty is forthcoming in Political
Studies.
C. FACTUAL INFORMATION
1. A key assumption underlying national
and international organisations' commitment to monitoring public
opinion is that public support for development assistance is fundamental
to maintaining, and importantly, increasing current levels of
aid. By way of example, the International Development Committee
in their inquiry call states: "Public support is essential
to an effective development policy and this could also be threatened
if the financial crises continues to affect the real economy.
As part of this inquiry, the committee will assess whether there
have been changes in public support for aid expenditure as a result
of the economic downturn, and in particular whether the public's
view on the UK Government's development work have changed".
In policy terms then, high levels of public support for development
aid, can be used as both an indicator to, and leverage against,
politicians who vote to allocate foreign aid against a number
of competing budget issues (Council of Europe, 2001; Burnell,
1998).
2. The assumed linkage between opinion and
policy is not without merit; a substantial literature has demonstrated
both the theoretical and empirical link between public opinion
and public policy (Monroe 1998; Hill & Hinton-Andersson 1995;
Stimson et al. 1995; Page and Shapiro 1983). Much of the
evidence, however, for a link between public opinion and policy
focuses predominantly public opinion and domestic policy
issues; the linkage between public opinion and foreign policy
has been harder to detect and unstable (Holsti 1992; 2004). This
is due, in part, to issue salience (ie the extent to which an
issue resonates with individuals) which has been shown to play
a key role in opinion formation and responsiveness (Burnstein
2003) and citizens' ability to process complex information and
reveal consistent opinions (Zaller 1992).
3. The assumed opinion-policy linkage has
not been subjected to robust empirical testing. In fact, evidence
produced by McDonnell (2001) shows a small, negative relationship
between opinion and aid. In other words, as support for aid increases,
aid allocations decrease.
4. Our analysis (see figure 1 and 2)
shows no evidence of a positive relationship between public opinion
and ODA (as a percentage of GNI) for OECD countries from 1990-2007.
We create two measures for public support. The first measure (figure
1) includes only those survey questions that ask explicitly about
support for development aid (PIPA) or the principle of development
aid (Eurobarometer). These questions are considered to be of greater
face validity—they measure what the researcher is intending
to measure, eg support for a particular policy of development
assistance—and therefore are the best indicators of public
support. Using this measure we find that public support averages
73.6% (standard deviation 12.6), however, there is a negative
correlation between support and ODA (r= -.04, p < .699, n=95).
This result is not statistically significant at the traditional
.05 level, which may be due to the small sample size (a result
of available data).
5. The second measure (figure 2) is more
inclusive, keeping the questions from our first measure as well
as those survey questions that tap concern for poverty in developing
countries more broadly. For example, the standard attitudinal
measure for support for development programs in the Eurobarometer
surveys ask, "In your opinion, is it very important, important,
not very important, or not at all important to help poor people
in countries in Africa, South America, Asia, etc. to develop",
and DFID's primary dependent variable asks respondents' level
of concern for poverty in developing countries (5-pt). Using this
measure, which we argue is a less valid measure of support for
development assistance (but more inclusive of general attitudes
towards poverty alleviation), we find pubic support averages 77.2%
(standard deviation= 13.1) and a small but statistically insignificant
relationship between support and ODA (r = .021, p < .81, n=
128). (See figure 1). In sum, both measures show essentially no
relationship between the two variables. The shift from a very
weak negative relationship to a weaker positive relationship is
not surprising given the change from a more restrictive measure
1, to a broader measure 2. In both cases, the slopes are essentially
flat indicating no relationship between the two variables.
6. Existing surveys on public opinion and
development assistance suffer from a number of problems that render
their findings suspect. We have examined the following representative
sample of surveys from which we offer the following critique:
(1) World Opinion on Addressing Hunger and Poverty (PIPA 2008);
(2) Public Attitudes towards Development (UK Department for International
Development 2008, 2005); (3) Europeans and Development Aid (European
Commission Eurobarometer 2007, 2005, 1999); and (4) Americans
on Foreign Aid and World Hunger (PIPA 2001). Our critique covers
four categories of interest: (1) the validity of measures of public
support for development assistance; (2) knowledge and awareness
of development; (3) perceptions of effectiveness and corruption,
and (4) motivations of support for development aid and concern
for poverty.
7. Validity 1. Existing surveys are
of poor validity, in other words they fail to measure the underlying
attitudes they aim to. To illustrate, the most commonly asked
question is whether respondents support the "principle of
providing aid to developing countries" (OECD, Eurobarometer).
From this and similar questions (eg respondents' level of concern
for poverty in developing countries) researchers conclude there
are high-levels of support for development assistance. However,
as Czaplinska (2007: 7) notes, questions that ask about helping
poor people and support for development aid are not equivalent.
Following on from this, interpreting high-levels of support from
such findings may be the result of measurement error, as respondents
may feel compelled to reply in the affirmative to the principle
of aid as a socially desirable response.
8. Validity 2. Survey instruments
need to do a better job of measuring absolute vs relative support
for development aid. Policymakers have assumed that consistently
high levels of support (>70%) readily translates into political
support, however there is no evidence that development assistance
as a policy issue is salient for respondents when compared to
competing policy issues. In measuring relative support, respondents
are asked to make reasoned trade-offs between competing policy
issues thus revealing relative policy preferences. For example,
when a PIPA (2001) survey asked respondents to register their
priorities, foreign aid is seen as less important that domestic
concerns. An overwhelming 84% agreed with the assertion that "taking
care of problems at home is more important than giving aid to
foreign countries" (PIPA 2001: 9).
9. Validity 3. Finally, research
has also shown that development aid/assistance can be interpreted
by respondents as at least three quite different things, namely
"1) attitudes toward humanitarian aid, 2) attitudes toward
ODA in relation to other government expenditures, and 3) views
to the effectiveness of different ODA activities" (Boas 2002:
2). McDonnell (2003) claims "public opinion does not even
see 'helping poor people' as 'development' because they conceive
aid as short term charity for humanitarian relief" (in McDonnell
2006, 5). She tests and finds support for the hypothesis that
public opinion has an impact on decisions on aid, only if "aid"
is specified and restricted to mean emergency aid. Evidence from
the UK has found that the public tends to associate the UK Government's
involvement in development as being disaster relief (42%) (DFID
2008: Q6).
10. Knowledge. Surveys of public
opinion and development are characterized by pervasive and persistent
low-levels of knowledge. Whilst this may be undesirable in light
of considerations of democratic competence and government responsiveness,
our criticism of low-levels of knowledge are targeted not towards
citizens of donor countries, but to researchers who fail to report
findings controlling for knowledge and indeed interest
levels. A second, but related criticism here is that Gilens (2001)
has shown that ignorance of policy specific information
leads citizens to hold views different then they might otherwise.
In short, policy specific information may yield different preferences
than general political knowledge, thus, it is not enough to be
interested or informed about politics generally, one must be able
to engage with policy specific knowledge in order to express genuine
attitudes and preferences.
11. Effectiveness. Evidence suggests
that support for development assistance is highly contingent on
perceptions of effectiveness, most especially corruption. In the
UK, for instance, 47% of respondents thought that aid was wasted
with 73% attributing this to corruption and 50% agreeing that
it was due to poor management and/or delivery (DFID 2008: 44).
Similarly, 53.8% of US respondents thought that aid money that
goes to poor countries ends up in the pockets of corrupt government
officials there (PIPA 2001: Q51). Though problems of corruption
appear to be reasonably well surveyed they are not analysed systematically
nor controlled for when reporting absolute levels of support.
12. Determinants of support 1. Whilst
there is a substantial literature on the determinants of state
motivations in giving aid, the determinants of individual-level
motivations is significantly under-researched. Two recent studies,
Paxton and Knack (2008) and van Heerde and Hudson (forthcoming
2009) have begun to examine what motivates individual-level support
for aid programmes or concern for poverty, but much more needs
to be done to understand the determinants of support particularly
as strategic communication(s) on development issues and the citizenship
education curriculum would be better targeted. This research agenda
needs to move beyond the traditional self-interest vs moral
imperative framework and better demarcate self-interest as direct
benefits (eg view of the world as a zero sum game or enlightened
self-interest (eg indirect benefits, positive sum game) in addition
to other explanations such as altruism (satisfaction gained by
others' utility (ie their utility is included in your preference
set), justice, cosmopolitanism (transnational identity or as global
citizens), religious beliefs, humanitarian or moral sympathies,
emotional responses, etc.
13. Determinants of support 2. Finally,
reporting public attitudes towards development aid should not
exist in isolation from measures of general attitudinal structures.
Or in other words, "polls rarely deliver information on underlying
attitudes and values and on preferable directions of aid. In addition,
polls usually neglect people's approach to policy issues going
beyond aid, for instance trade, debt and immigration policy"
(Czaplinska 2007: 14). Here, support for development aid may co-vary
with attitudes on domestic welfare redistribution (Noel &
Therein 2002), position on a left-right ideological scale, perceptions
of government efficiency, or trust in political institutions.
On the latter, a recent DFID survey (2008) poses a general question
on respondents' trust in government to act in the country's best
interest. Knowing something about respondents' wider attitudinal
structures may provide a more robust interpretation of support
for development aid, and importantly, link existing "publics"
of support on domestic policy issues with "publics"
of support for development aid.
D. RECOMMENDATIONS
1. There dangers to DFID's mission of poverty
alleviation in rendering future policy decisions on development
aid contingent upon results from existing surveys of public support
for development aid programmes. Whilst democratic responsiveness
is key to effective democratic governance, current measures of
public support are subject to extreme measurement error and are
of poor validity. Consequently, whilst we might expect a decline
in support as a result of the economic downturn, any such decline
should be interpreted with serious caution as existing (high)
levels of support may also be dubious or the function of measurement
error.
2. During an economic downturn, public opinion
tends to favour spending on domestic policy issues/problems relative
to foreign aid expenditures. However, this opinion is no different
than under normal economic conditions. Individuals have consistent
preferences for spending at home than abroad. The Government should
not view hard economic times as being different vis-à-vis
policy preferences on development assistance than normal economic
conditions.
3. It is crucial that knowledge questions
are integrated into both survey instrument design and into
the analysis of responses; respondent knowledge questions are
fundamental to evaluating the robustness of public support. Moreover,
issue salience should be controlled for when determining levels
of support. The policy implication of this is the UK government
needs to be able to identify those constituencies whose support
is robust (either positive or negative) and those whose support
variable or non-existent.
4. Let the status quo persevere: the UK
government should not pull back from its 2013 commitment
to meeting the 0.7% target. Public opinion may be a vital
ingredient in determining aid levels and programmatic activity,
but it is currently not measured in such a way that can be considered
valid and reliable.
5. We recommend significant modifications
to DIFD's (and others) monitoring of public attitudes in order
to reveal robust support for development assistance. Understanding
what drives individual-level motivations, especially in young-people,
it vital for tailoring curriculum design and strategic communications
to strengthen support for development assistance.
REFERENCES
Boas, M. (2002). Public attitudes to aid in
Japan and Norway. Working Paper 2002/02, Centre for Development
and Environment, University of Oslo.
Czaplinska, A. (2007). Building support for
development cooperation. ICEI: Instituto Complutense de Estudios
Internacionales: PP 02/07. http://www.ucm.es/info/icei/pdf/PP%2002-07.pdf
[Accessed 08 February 2009.].
Council of Europe (2001). "0.7%: When and
How?" http://www.coe.int/t/e/northsouth_centre/resources_centre/1_the_interdependent/Terraviva_GB101.pdf.
The North-South Centre. Strasbourg.
Department for International Development (DFID).
(2005). Public attitudes on development; beliefs and attitudes
concerning poverty in developing countries. London.
van Heerde, J., & Hudson, D. (forthcoming).
The righteous considereth the cause of the poor? Public attitudes
towards poverty in developing countries. Political Studies.
Holsti, O. (1994). Public opinion and American
foreign policy. Ann Arbor: The University of Michigan Press.
Holsti, (1992). "Public Opinion and Foreign
policy: Challenges to the Almond-Lippman Consensus", International
Studies Quarterly, 36 (4), 439-466.
Mc Donnell, I. (2001). The state of public support
for development co-operation. Terra Viva Dossier. North-South
Centre/Inter Press Service. www.nscentre.org/tvmonthly/. [Accessed
December 2008].
McDonnell, I., H. Lecomte, & L. Wegimont,
eds. (2003). Public Opinion and the Fight Against Poverty.
OECD/North-South Centre. Paris: France.
Monroe, A.D. (1998). Public opinion and public
policy. Public Opinion Quarterly 62(6), 6-28.
OECD Development Centre. (2008). Public opinion
polls about development.
http://www.oecd.org/document/6/0,3343,en_2649_34101_39295366_1_1_1_1,00.html.
[Accessed 9 December 2008].
Page, B., & Shapiro, R. (1983). Effects
of public opinion on policy. American Political Science Review
77(1), 175-190.
Program on International Policy Attitudes (PIPA).
(2001). Americans on foreign aid and world hunger: A study of
US public attitudes. Available at: http://www.pipa.org/OnlineReports/ForeignAid/
ForeignAid_Feb01/ForeignAid_Feb01_rpt.pdf [Accessed
27 August 2008].
Smillie, I. (1996). Mixed messages: Public opinion
and development assistance in the 1990s. In C. Foy and H. Helmich,
eds., Public support for international development. Development
Centre OECD: Paris, 27-54.
Stern, M. (1998). Development aid: What the
public thinks? United Nations Development Programme, Office
of Development Studies. New York: New York.
WorldPublicOpinion.org (2008). World Opinion
on Addressing Hunger and Poverty, Publics in Developed Countries
Ready to Contribute Funds Necessary to Cut Hunger in Half By 2015.
PIPA: University of Maryland. Available at: http://www.worldpublicopinion.org/pipa/articles/btdevelopmentaidra/554.php?nid=&id=&pnt=554&lb=btda.
[Accessed 12 November 2008].
Zaller, J. (1992). The nature and origins
of mass opinion. Cambridge: Cambridge University Press.
Figure 1. Correlation of Public Support
for the (principle of) Development Aid and ODA, 1990-2007 OECD
Countries
Figure 2. Correlation between Public Support
for Development Aid/Concern for Poverty and ODA, 1990-2007 OECD
Countries
| 
