VoxEU Column Monetary Policy

Central bank independence and transparency: Not just cheap talk (Part 2)

Theories arguing that independent, transparent central banks fight inflation better are widely accepted, but the evidence backing them is surprisingly scarce. This column presents new empirical estimates suggesting a payoff to central bank independence and transparency.

Using the updated measures of central bank independence and transparency that we detailed in our first column, we sought to investigate what effects these aspects of central bank governance might have on economic performance.1

While the theoretical case for central bank independence appears to have been accepted, empirical studies have found surprisingly limited evidence of independence delivering its promised anti-inflation benefits in practice. Hence, while the earliest studies of independence focusing on a fairly narrow subset of industrial countries delivered this result, later studies covering a wider set of developing and industrial countries found more equivocal results.

Independence and inflation

We estimated the relationship between independence and inflation for the 1980s and the current period using the original data from Cukierman, Webb, and Neyapti (1992) and our new measure and confirmed the puzzlingly weak or non-existent relationship between independence and inflation found in the literature (see the first two columns of Table 1). We measured central bank autonomy using both our de jure index (CBI) and a de facto measure, the turnover rate for the central bank governor over a ten-year period (TURNOVER). A higher turnover rate suggests lower de facto independence. Controls included real GDP per capita (GDP), trade openness (OPEN), and the de facto exchange rate regime (REGIME, an indicator variable that increases in the degree of exchange rate flexibility). These controls are expressed as five-year averages, measured over 1987-91 (period 0) and 2002-06 (period 1).

However, when we revisited this relationship exploiting the time dimension of our data we found robust evidence for the negative relationship between central bank independence and inflation predicted by theory, for both the de facto and de jure measures. The added time dimension allows us to employ a first differenced specification, eliminating unobservable and omitted country fixed effects that are likely to be correlated with the levels of independence. The results, reported in column 3 of Table 1, are quantitatively significant, suggesting, for instance, that the average increase in CBI (around .25) is associated with a drop in inflation of more than 5 percentage points compared to a situation where independence does not increase at all.

We repeated this exercise looking at the individual subcomponents of the index (column 4 of Table 1) and found that having a single, well-specified, ideally numerical inflation or price level target for the central bank is the aspect of overall independence most likely to deliver low inflation. Hence, central banks’ legal objectives do not appear to be cheap talk – they may help the central bank commit to its anti-inflationary goals.

Table 1 Inflation and central bank independence

  (1) (2) (3) (4)
Dependent variable: Inflation, period 0 Inflation, period 1
Change in inflation



TURNOVER 0.61***
GDP 4.01
OPEN -1.06***
REGIME -0.00
CBI1       -0.01
CBI2       -0.02
CBI3       -0.19**
CBI4       -0.05
Obs. 56 56 56 56
R2 0.44 0.18 0.31 0.36

Notes: Ordinary least squares estimation with robust standard errors reported in parentheses. Equations (1) and (2) estimated with explanatory variables in levels; equations (3) and (4) estimated with explanatory variables in first differences. Constants included but not reported. */**/*** denote significance at the 10, 5, and 1 percent levels, respectively. OPEN divided by 1,000 to aid presentation of results. The dependent variable is the inflation tax transform π/(1+π).


We also attempted to control for endogenous reform by instrumenting for the change in the independence index using two measures of overall governance for each country: the rule of law and voice and accountability (both taken from the World Bank’s Governance Matters data). With instrumentation, the estimated effect of independence is somewhat stronger than that estimated via ordinary least squares, and remains negative.

What are the effects of transparency?

In theory, increased central bank transparency can have a number of implications for macroeconomic variables, but these tend to be rather model-specific – general lessons are hard to tease out. Transparency tends to be beneficial when information asymmetries are themselves the cause of inefficiencies in the economy but can be costly in a second-best environment where the central bank is able to offset other inefficiencies by exploiting its informational advantage. Ultimately, the question of whether central bank transparency delivers tangible benefits is an empirical one.

To shed some light on this question we focused on private sector inflation forecasts, since they seems to us the most fertile ground for assessing the impact of different transparency practices. In particular, we wanted to test whether greater central bank transparency leads to greater use of public (as opposed to private information) by the private sector. The private sector’s forecasts of inflation are the central bank’s ultimate objective and the principal object of its communication strategy.

As a measure of the extent of private information versus public information relied upon by private sector forecasters, we used the ratio of the variance of the private sector forecasts for inflation in the following year to the mean square error of the forecasts.2 This measure lies on the unit interval because the mean square error is by definition the sum of the forecast variance and the square of the bias. Improvements in transparency arising from more accurate public signals should reduce the ratio, as when public information is more accurate, all forecasters should place greater weight on it compared to their private information, leading to less variance across forecasters.

We used the ratio of the variance to the mean squared error of private sector inflation forecasts as the dependent variable in a regression and controlled for trade openness (OPEN1) and governance (regulatory quality, RQ1, from the World Bank’s Governance Matters data). Other macroeconomic controls were not found to have any effect, so we dropped them.
We found robust evidence for the predicted negative correlation, indicating that greater transparency is associated with more accurate private sector forecasts. The estimated effect is both quantitatively and statistically significant (see column 1 of Table 2). When we estimated the relationship separately for the five components of our transparency measure, we found that the most significant effects come from those components of our overall index most closely related to the issuance of the central bank’s economic forecasts, strongly suggesting that the estimated effect is not merely an artefact of the data (columns 2-7).

Table 2. Private information and central bank transparency

Dependent variable: Ratio of variance to mean squared error of private sector inflation forecasts
  (1) (2) (3) (4) (5) (6) (7)
TRANS1 -0.32**
OPEN1 -0.85**
RQ1 0.12***
T1.1   -0.08
T2.1     -0.27**
T3.1       -0.10
T4.1         0.02
T5.1           -0.20***
Obs. 28 28 28 28 28 28 28
R2 0.27 0.13 0.31 0.16 0.10 0.31 0.47

Notes: Ordinary least squares estimation with robust standard errors reported in parentheses. Constants included but not reported. */**/*** denote significance at the 10, 5, and 1 percent levels, respectively. OPEN1 divided by 1,000 to aid presentation of results.


The baseline regression results indicate that a one standard deviation increase in the transparency score (around .22) is associated with a seven percentage point decline in the variance to mean squared error ratio. This could point to a quite substantial increase in the accuracy of public information associated with an increase in transparency. For instance, if we assume that public and private information are initially equally accurate (in the sense of having an equal variance) and that the variance to mean squared error ratio were initially at its mean value (0.26), then the .07 decrease in the ratio corresponds to an increase in the relative accuracy of public information of around 70 percent.

We hope that our data (publicly available for download from the IMF website) will be widely used to test other hypotheses in the area of central bank governance.


Crowe, C., Meade, E. E., 2008. Central Bank Independence and Transparency: Evolution and Effectiveness. European Journal of Political Economy, Forthcoming. Data: http://www.imf.org/external/pubs/ft/wp/2008/data/wp08119.zip
Cukierman, A., Webb, S.B., Neyapti, B., 1992. Measuring the Independence of Central Banks and Its Effect on Policy Outcomes. The World Bank Economic Review 6, 353--398.



1 The views expressed herein are those of the authors and should not be attributed to the IMF, its Executive Board, or its management.
2 We used the next year’s actual inflation outturn as the measure of the true outcome that the forecasters are attempting to forecast.


1,575 Reads