Why the Economy?

1
Why the Economy?

• Raymond Duch
• University of Houston
• Randy Stevenson
• Rice University

2
Cross-national Studies of Economic Voting
3
(No Transcript)
4

• At the individual level, limited efforts to
  understand cross-national variation.
• Lewis-Beck 1988
• Paldam 1991
• Nonetheless some interesting differences in model
  results cross-nationally

5

• At the aggregate level, considerable
  cross-national variation in economic voting
  results.
• Clarity of responsibility is one effort to
  help understand but limited.

6
Research Puzzle

• We suspect individual-level economic voting
  models vary significantly cross-nationally.
• But current published modeling efforts are not
  rich enough to draw meaningful conclusions.

7
Our Insights

• Economic evaluations are attitudes information
  processing, cueing and cognitive psychology
  literature.
• Measure extent to which economic information is
  mediated.

8
(No Transcript)
9
The Research Project
10
(No Transcript)
11
Mediated Economic Cues

• Citizens employ cues/information short-cuts in
  forming economic assessments
• Luppia McCubbins argument
• Nevertheless information gathering costs matter

12

• Political actors, elites, media attempt to shape
  these cues or packaging of messages
• Zaller on the strategic manipulation of media by
  political actors
• Iyengar on framing

13

• Distorted messages regarding economy?
• Some evidence from analyses of different media
• Negativity bias
• Human interest bias

14
Distortions depend on

• Information gathering costs
• Cross-national diversity of media
• Institutional contexts

15
Implications for EV Models?

• Economic evaluations are based on distorted
  information
• Seriously questions the democratic
  accountability theory of EV

16
Null hypothesis

• Economic perceptions and actual economic outcomes
  track each other in a regular fashion
• The series are co-integrated and can be modeled
  as an ECM

17
Mediated cues argument

• There is distortion in mass assessments of
  economic outcomes.
• This distortion is a function of mediated
  messages regarding the economy (media,
  politicians, etc.)

18
Hypothesis

• Distortion evidenced by poor ECM fit.
• Mediation implies degree of distortion correlated
  with cost of information
• High for unemployment
• Low for inflation
• Mediation implies asymmetric effects

19

• Mediation also implies considerable
  cross-national variation in degrees of distortion
• Mediation implies distortions in both mass and
  elite economic assessments

20
Data

• Mass publics economic assessments
• Monthly consumer confidence surveys conducted by
  the European Commission in all member countries
  1986-2000
• Elites economic assessments
• Monthly business confidence surveys from European
  Commission 1967-2000

21

• Economic indicators
• Quarterly GDP growth from OECD (Palmer and
  Whitten) CPI and unemployment figures from The
  Economist
• All variables are transformed into standardized z
  scores

22
Modeling Economic Attitudes
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
The error correction model
DProInfit ai - d ProInfit-1 - t1INFit
b1DINFit1 eit
28

• d
• t-test on d indicates significance of
  co-integration

29

• t1
• long term multiplier for the effect of a change
  in the real economy on economic expectations
• Expectation is that t1 1

30
Results
31
(No Transcript)
32
CPI and Price Expectations

• d (ECM) parameter significant in all but one
  country (Italy)
• The long term relationship between CPI and Price
  Expectations captured by t1 is significant and
  close to 1 as expected
• Note the asymmetry -- t1 is larger when inflation
  increasing

33
(No Transcript)
34
Unemployment and Unemployment Expectations

• d (ECM) parameter significant is not significant
• Suggesting the two series are not co-integrated.

35
(No Transcript)
36
General Economy GDP

• d (ECM) parameter significant in all countries
• The long term relationship between general
  economy and GDP captured by t1 is significant and
  close to 1 in vast majority of case
• Asymmetry in t1 is less consistent but evident

37
Cross national variation

• Note there is cross-national variation in
• The magnitude of t1
• The incidence of asymmetry

38
Conclusions

• Perceptions of the economy and the real economy
  are in many cases co-integrated and seem to fit
  an error correction process
• There is evidence of distortion

39

• There is also evidence that distortion results
  from mediated cues
• Economic outcomes that are costly to monitor
  (unemployment) are perceived less well by the
  mass public than easily monitored outcomes (CPI)
• Asymmetry in reactions to positive and negative
  economic news

40
Cross-national variation

• Long-term equilibrium relationship between series
• Asymmetry varies by nation
• Suggests role of mediated economic information

41
The economy, economic attitudes and political
preferences
42
What Drives Popularity Series?

• Subjective assessments of the economy which we
  think are shaped from mediated representations of
  the economy?
• Or, the real economy?

43
(No Transcript)
44
(No Transcript)
45
(No Transcript)