Chapter 12 Quiz: Monetary Policy and Data/Parameter Uncertainties
1. Additive uncertainty refers to the central bank's uncertainty about the parameters of the economic model.
False. Additive uncertainty, also known as data uncertainty, refers to uncertainty about the true state of the economy due to random shocks and data measurement errors. The central bank is assumed to know the model's parameters. Uncertainty about the parameters is known as parameter (or multiplicative) uncertainty. (Source: EC3115 Subject Guide, Section 12.1, 12.7)
2. According to Aruoba (2008), 'well-behaved' data revisions should have a non-zero mean, indicating a systematic bias in initial data releases.
False. Aruoba (2008) states that one of the properties of 'well-behaved' revisions is that they should have a mean of zero. A non-zero mean implies that the initial announcements are biased estimates of the final values. (Source: Aruoba, 2008, p. 320; EC3115 Subject Guide, Section 12.7)
3. 'News' in data revisions refers to situations where the initial data announcement is an observation of the final series measured with error, and the revision is correlated with available data at the time of the initial announcement.
False. This describes 'noise'. 'News' is when the initial announcement is an efficient forecast reflecting all available information. Subsequent revisions ('news') are due to new information and should be uncorrelated with data available at the time of the initial announcement. (Source: Aruoba, 2008, p. 327)
4. The 'noise' hypothesis of data revisions suggests that the initial data release is an optimal forecast of the true value.
False. The 'noise' hypothesis posits that the initial announcement is the true value measured with error \(y_t^{t+1} = y_t^f - r_t^f\). The 'news' hypothesis posits that the initial announcement is an optimal forecast. (Source: Aruoba, 2008, p. 327)
5. Aruoba (2008) finds strong empirical evidence that macroeconomic data revisions in the United States are generally 'well-behaved'.
False. Aruoba's central finding is that data revisions are *not* well-behaved. He finds that revisions often have non-zero means (are biased), are large compared to the original variables (high noise-to-signal ratio), and are predictable using information available at the time of the initial announcement. (Source: Aruoba, 2008, p. 319)
6. Additive shocks in the IS-LM framework, such as unexpected changes in consumer tastes or financial market sentiment, create data uncertainty for the policymaker.
True. In Poole's (1970) model, these random shocks (denoted by \(\epsilon\) for the IS curve and \(\eta\) for the LM curve) are the source of additive uncertainty. The central bank cannot observe these shocks directly, only their effects on output and interest rates, creating uncertainty about the true position of the IS and LM curves. (Source: EC3115 Subject Guide, Section 12.7)
7. If data revisions are pure 'news', then they should be predictable.
False. If data revisions are 'news', they represent the arrival of new, unforeseen information. By definition, they should be unpredictable based on the information set available at the time of the initial forecast. Predictability is a characteristic of 'noise'. (Source: EC3115 Subject Guide, Section 12.7; Aruoba, 2008, p. 327)
8. Aruoba (2008) concludes that the 'news' versus 'noise' framework is robust and most variables can be clearly classified as one or the other.
False. Aruoba (2008) finds that for most variables, both the news and noise hypotheses are rejected, leading to an 'ambiguous conclusion'. This suggests the simple dichotomy is insufficient to describe the complex nature of data revisions. (Source: Aruoba, 2008, p. 328)
9. The final revision of a data series is defined as the difference between the very first announcement and the latest available data point years later.
True. In the literature on real-time data, the 'final' revision is typically defined as the difference between the initial announcement \(y_t^{t+1}\) and the 'true' or final value \(y_t^f\), which is taken to be the value from the latest available data vintage. (Source: Aruoba, 2008, p. 323; EC3115 Subject Guide, Section 12.7)
10. A high 'noise-to-signal' ratio in data revisions, as found by Aruoba (2008), suggests that the magnitude of data revisions is small relative to the variability of the actual economic variable.
False. A high noise-to-signal ratio (variance of the revision divided by the variance of the final value) indicates that the revisions are large and volatile relative to the variable itself. Aruoba (2008) finds these ratios to be quite large, suggesting revisions are a significant source of uncertainty. (Source: Aruoba, 2008, p. 325)
11. In the Poole (1970) model, if the economy is primarily subject to shocks in the goods market (IS shocks), the central bank should target the money supply to minimize output fluctuations.
True. When IS shocks dominate, holding the money supply constant allows the interest rate to act as an automatic stabilizer. A positive IS shock pushes output up, but this also increases money demand, raising interest rates and dampening the initial shock. Targeting the interest rate would require increasing the money supply, which would fully accommodate the IS shock and lead to greater output volatility. (Source: Poole, 1970; EC3115 Subject Guide, Section 12.7; Unit L Lectures, Slides 5-7)
12. If the money market is unstable and subject to shocks (LM shocks), Poole's analysis suggests that targeting the interest rate is the optimal policy for output stabilization.
True. When LM shocks (e.g., shifts in money demand) dominate, fixing the interest rate completely insulates the goods market and output from these shocks. The central bank simply adjusts the money supply to meet the fluctuating demand at the target interest rate, keeping output stable. Targeting the money supply would transmit these financial shocks to the interest rate and then to the real economy, causing output volatility. (Source: Poole, 1970; EC3115 Subject Guide, Section 12.7; Unit L Lectures, Slides 11-13)
13. When a central bank targets the interest rate, it gives up control over the money supply.
True. A central bank can choose to control a monetary aggregate (like M1) or an interest rate, but not both simultaneously. If it commits to a specific interest rate target, it must be willing to supply whatever quantity of money is demanded at that rate. (Source: Poole, 1970; Unit L Lectures, Slide 4)
14. According to Poole's analysis, if both IS and LM shocks are present, the optimal policy is always to target the money supply.
False. The optimal policy depends on the relative variances of the IS and LM shocks and the structural parameters of the model. The choice is a trade-off. The more dominant IS shocks are, the more attractive money supply targeting becomes. The more dominant LM shocks are, the more attractive interest rate targeting becomes. (Source: Poole, 1970; EC3115 Subject Guide, Section 12.7)
15. An interest rate targeting policy completely eliminates output fluctuations when all shocks are to the LM curve.
True. With pure LM shocks, the IS curve is stable. By holding the interest rate constant, the central bank ensures the equilibrium point on the IS curve does not change. All instability from the money market is absorbed by changes in the money supply, leaving output unaffected. The output variance is zero. (Source: EC3115 Subject Guide, Table 12.1; Unit L Lectures, Slide 13)
16. A money supply targeting policy completely eliminates output fluctuations when all shocks are to the IS curve.
False. With a money supply target, the LM curve is fixed. When an IS shock occurs, the IS curve shifts along the fixed LM curve. The interest rate changes, which partially offsets the shock, but output still deviates from its target. Output variance is reduced compared to an interest rate target, but not eliminated. (Source: EC3115 Subject Guide, Section 12.7; Unit L Lectures, Slide 5)
17. The widespread shift by central banks from monetary targeting to interest rate targeting in the 1980s can be explained by increasing instability in the goods market (IS shocks).
False. This shift is typically explained by increasing instability in the money market (LM shocks), largely due to financial innovation and deregulation. As money demand became more volatile and unpredictable, targeting the money supply created excessive interest rate and output volatility, making interest rate targeting the more prudent strategy. (Source: Unit L Lectures, Slide 15)
18. In Poole's model, the variance of output under an interest rate target is given by \(\sigma_\epsilon^2\), where \(\sigma_\epsilon^2\) is the variance of the IS shock.
True. When the interest rate is fixed, the central bank must adjust the money supply to accommodate any shocks. For an IS shock \(\epsilon\), the central bank's policy effectively makes the LM curve horizontal, so the full impact of the shock is felt on output. Thus, the variance of output is simply the variance of the shock itself. (Source: EC3115 Subject Guide, Table 12.1)
19. In Poole's model, the variance of output under a money supply target depends on the variances of both IS and LM shocks.
True. The formula for output variance under money supply targeting is \( \frac{1}{(d+be)^2} (d^2\sigma_\epsilon^2 + b^2\sigma_\eta^2) \). This shows that output is affected by shocks from both the goods market (\(\sigma_\epsilon^2\)) and the money market (\(\sigma_\eta^2\)). (Source: EC3115 Subject Guide, eq. 12.5)
20. Poole's (1970) analysis assumes that the central bank knows the true parameters of the IS and LM curves.
True. Poole's model deals with additive (data) uncertainty, where the structure of the economy is known, but it is hit by unobservable random shocks. It does not address parameter uncertainty, which was later analyzed by Brainard (1967). (Source: EC3115 Subject Guide, Section 12.7)
21. Parameter uncertainty arises when the central bank is unsure about the magnitude of the effect its policy actions have on the economy.
True. Parameter uncertainty (or multiplicative uncertainty) refers to the central bank's lack of precise knowledge about the coefficients in its model of the economy, such as the interest rate elasticity of investment or the policy multiplier. (Source: Brainard, 1967, p. 412; EC3115 Subject Guide, Section 12.8)
22. The presence of parameter uncertainty reinforces the principle of certainty equivalence.
False. The presence of parameter uncertainty is precisely the condition that *breaks* the certainty equivalence principle. Because the variance of the outcome depends on the policy action taken, the policymaker must consider this variance and will not simply act on expected values as if they were certain. (Source: Brainard, 1967, p. 413; EC3115 Subject Guide, Section 12.8.1)
23. According to Brainard (1967), when there is uncertainty about a policy instrument's multiplier, the central bank should use that instrument more aggressively to ensure its target is met.
False. Brainard's key result is that parameter uncertainty leads to caution, gradualism, or 'attenuation'. Policymakers should use the instrument *less* aggressively than they would under certainty. This is because aggressive actions amplify the variance of the outcome when the multiplier is uncertain. (Source: Brainard, 1967, p. 415; EC3115 Subject Guide, Section 12.8)
24. In Brainard's model, if the policy multiplier is uncertain, the optimal policy is to only partially close the expected gap between the target variable and its desired value.
True. The optimal policy trades off closing the mean gap against the cost of increased variance. The optimal policy setting is \(X = \frac{\bar{g}y^*}{\sigma_g^2 + \bar{g}^2}\), which is less than the certainty equivalent policy of \(X = y^*/\bar{g}\). This implies the expected output gap is not fully closed. (Source: Brainard, 1967; EC3115 Subject Guide, eq. 12.10)
25. 'Brainard conservatism' refers to the idea that policymakers should react more forcefully to shocks when they are uncertain about the economy's structure.
False. 'Brainard conservatism' or 'gradualism' is the principle that parameter uncertainty should lead to a more cautious and less aggressive policy response. The central bank 'discounts' its policy actions to avoid inducing too much volatility. (Source: EC3115 Subject Guide, Section 12.8.1)
26. If a central bank has two policy instruments with uncertain multipliers to hit one target, Brainard's analysis suggests it should use only the instrument with the least uncertainty.
False. Brainard's analysis shows that it is generally optimal to use a portfolio of all available instruments, even if one is clearly 'better' (less uncertain) than the other. This is a diversification principle: combining instruments can create a 'policy portfolio' with a lower overall variance of impact than any single instrument used alone. (Source: Brainard, 1967, p. 418-419)
27. In the Brainard model, an increase in the degree of parameter uncertainty (a higher variance of the policy multiplier) leads to a more cautious policy response.
True. In the optimal policy formula \(X = \frac{\bar{g}y^*}{\sigma_g^2 + \bar{g}^2}\), a larger variance \(\sigma_g^2\) increases the denominator. This reduces the optimal setting for the policy instrument \(X\), meaning the policy becomes more cautious or attenuated. (Source: EC3115 Subject Guide, Section 12.8; Unit L Lectures, Slide 28)
28. Parameter uncertainty is also known as multiplicative uncertainty because the uncertain parameter multiplies the policy instrument, meaning the variance of the outcome depends on the level of the instrument.
True. In an equation like \(y = gX\), the uncertain parameter \(g\) multiplies the instrument \(X\). The variance of output is \(Var(y) = X^2 Var(g) = X^2 \sigma_g^2\). This shows that the variance of the outcome is a function of the policy choice, which is the key feature of multiplicative (parameter) uncertainty. (Source: EC3115 Subject Guide, Section 12.1, 12.8)
29. The 'attenuation principle' suggests that the more uncertain a policy's effects are, the larger the policy action should be.
False. The attenuation principle, derived from Brainard's work, states the opposite: the more uncertain a policy's effects are (i.e., the higher the parameter uncertainty), the more cautiously it should be used. The policy action should be attenuated or scaled back. (Source: Brainard, 1967)
30. When policy effects are uncertain, it can be optimal to use an instrument in the 'wrong' direction (e.g., raising rates when a rate cut seems needed) to reduce overall risk.
True. Brainard (1967) shows that if the impacts of two policy instruments are correlated, or if an instrument's impact is correlated with an exogenous disturbance, it can be optimal to use one instrument in a counter-intuitive direction. This is done not to move the mean of the target in that direction, but to hedge against risk, reducing the overall variance of the outcome. (Source: Brainard, 1967, p. 416, 420)
31. The certainty equivalence principle states that policymakers should act based on the expected values of uncertain variables as if they were certain.
True. This principle, developed by Theil and Simon, implies that for a quadratic objective function and a linear model with only additive uncertainty, the optimal policy is to replace all random variables with their unconditional means and solve the resulting deterministic problem. (Source: Brainard, 1967, p. 413; EC3115 Subject Guide, Section 12.3)
32. Certainty equivalence holds in a model with parameter (multiplicative) uncertainty.
False. Certainty equivalence breaks down under parameter uncertainty. This is because the variance of the target variable becomes a function of the policy instrument, so the policymaker cannot ignore this uncertainty and must trade off mean and variance. (Source: EC3115 Subject Guide, Section 12.8.1)
33. If a central bank's loss function is quadratic and its model of the economy is linear, certainty equivalence will always apply.
False. Certainty equivalence requires not only a quadratic loss function and linear model, but also that the uncertainty is purely additive (i.e., no parameter uncertainty). If there is parameter uncertainty, certainty equivalence does not hold. (Source: Brainard, 1967, p. 413)
34. Under certainty equivalence, the optimal policy is not affected by the variance of the additive shocks.
True. With additive uncertainty, the variance of the target variable is independent of the policy action (e.g., in \(y = aP + u\), \(Var(y) = Var(u)\) if \(a\) is known). Since the policymaker cannot influence this variance, they focus solely on setting the mean of \(y\) to its target, ignoring the variance of the shock \(u\). (Source: EC3115 Subject Guide, Section 12.8.1)
35. The concept of certainty equivalence provides a justification for policy gradualism.
False. Certainty equivalence implies that policymakers should move immediately and fully to the setting that is expected to achieve their goal. Policy gradualism or caution is a departure from certainty equivalence, and it is justified by the presence of parameter uncertainty. (Source: EC3115 Subject Guide, Section 12.9)
36. In a New Keynesian model with only additive shocks (e.g., \(\pi_t = a\pi_{t-1} - bi_t + \epsilon_t\) where \(a\) and \(b\) are known), the optimal interest rate rule is the same as it would be if there were no shocks.
True. This is a direct application of the certainty equivalence principle. Since the shock \(\epsilon_t\) is additive and its variance is not affected by the policy choice \(i_t\), the central bank sets the interest rate to target the expected value of inflation, ignoring the shock term. The resulting rule is \(i_t = (a\pi_{t-1} - \pi^*)/b\), which is the same as in a deterministic world. (Source: EC3115 Subject Guide, Section 12.8.1)
37. The reason certainty equivalence fails under parameter uncertainty is that the policymaker's actions can alter the variance of the target variable.
True. With parameter uncertainty (e.g., \(y = gX\) where \(g\) is random), the variance of the outcome is \(X^2\sigma_g^2\). The policy instrument \(X\) now appears in the variance term. The policymaker must therefore consider how their actions affect not just the expected outcome, but also its volatility. (Source: Brainard, 1967, p. 413)
38. A policymaker who follows certainty equivalence in a world with parameter uncertainty will behave optimally.
False. Such a policymaker would behave sub-optimally. By ignoring the effect of their actions on outcome variance, they would likely choose a policy that is too aggressive, leading to excessive volatility and a lower expected utility (higher loss). (Source: Brainard, 1967, p. 415)
39. Certainty equivalence is a practical rule for policymaking because real-world uncertainty is almost exclusively additive.
False. This is doubly false. First, real-world uncertainty involves significant parameter uncertainty, not just additive shocks. Second, because of this, certainty equivalence is often *not* a practical rule and can be misleading. The prevalence of policy gradualism suggests that real-world policymakers implicitly reject certainty equivalence. (Source: EC3115 Subject Guide, Section 12.1)
40. The separation principle, where the policymaker first estimates the state of the system and then optimizes policy based on that estimate, is a feature of certainty equivalence.
True. Certainty equivalence allows for a two-stage process. First, form the best possible estimate (the conditional expectation) of the state of the world. Second, solve the policy problem as if this estimate were the true state of the world. This separation is not optimal when parameter uncertainty is present. (Source: Brainard, 1967)
41. If IS shocks become more volatile relative to LM shocks, a central bank following Poole's (1970) logic should consider shifting from interest rate targeting towards money supply targeting.
True. Poole's analysis shows that money supply targeting is more effective at stabilizing output against IS shocks. Therefore, an increase in the relative importance of IS shocks makes money supply targeting a more attractive strategy compared to interest rate targeting. (Source: Poole, 1970; EC3115 Subject Guide, Section 12.7)
42. Brainard's (1967) model of parameter uncertainty provides a theoretical justification for the 'cold turkey' approach to disinflation.
False. Brainard's model justifies the opposite: a 'gradualist' approach. The 'cold turkey' approach (a sharp, aggressive policy change) would be optimal under certainty equivalence, but with parameter uncertainty, such a policy would create excessive and undesirable volatility. (Source: EC3115 Subject Guide, Section 12.8.1)
43. Aruoba's (2008) finding that data revisions are predictable ('noise') implies that initial data releases are efficient forecasts.
False. Predictability of revisions implies that the initial data releases are *not* efficient forecasts. An efficient forecast should incorporate all available information, and its errors (the subsequent revisions) should be unpredictable. (Source: Aruoba, 2008, p. 327)
44. In the Poole model, the choice of instrument becomes irrelevant if there are no shocks to either the IS or LM curves.
True. In a deterministic world with no shocks, the IS and LM curves are perfectly stable. The central bank can choose a money supply (M*) and interest rate (i*) pair that lies on the LM curve and achieves the output target. There is no trade-off to consider. (Source: Unit L Lectures, Slide 3)
45. The policy trade-off in Brainard's model is between achieving the mean target and minimizing the variance of the outcome.
True. Due to parameter uncertainty, a more aggressive policy to close the gap in the mean of the target variable also increases the variance of the target variable. The optimal policy represents a trade-off between these two objectives, as illustrated by the tangency between the policy constraint and the policymaker's indifference curve. (Source: Unit L Lectures, Slides 21, 29-32)
46. If the central bank's policy multiplier has a coefficient of variation of 0.5, Brainard's model suggests the bank should aim to close 100% of the expected output gap.
False. Brainard's formula for the optimal policy is P* = g / (1 + V^2), where V is the coefficient of variation. If V = 0.5, then V^2 = 0.25. The optimal policy would be to close 1 / (1 + 0.25) = 1 / 1.25 = 80% of the gap, not 100%. (Source: Brainard, 1967, p. 415)
47. An interest rate peg (targeting) makes the effective LM curve horizontal.
True. By committing to a fixed interest rate, the central bank agrees to supply any amount of money demanded at that rate. This makes the LM curve perfectly elastic (horizontal) at the target interest rate. (Source: Unit L Lectures, Slide 7)
48. Aruoba (2008) finds that the unconditional mean of final revisions for most major US macroeconomic variables is statistically indistinguishable from zero.
False. A key finding of Aruoba (2008) is that the means of final revisions for a majority of the 19 variables studied are positive and statistically significant, indicating that initial announcements are systematically biased downwards. (Source: Aruoba, 2008, p. 324)
49. In Brainard's framework, if two policy instruments are available, the optimal policy mix depends only on the variance of their individual impacts, not the covariance between them.
False. The optimal policy portfolio explicitly depends on the covariance (or correlation) between the impacts of the instruments. A negative correlation is particularly valuable, as it allows for diversification that can significantly reduce the overall variance of the policy package. The formula for the optimal mix includes the covariance term \(\rho_{12}\sigma_{a1}\sigma_{a2}\). (Source: Brainard, 1967, p. 419)
50. The distinction between 'news' and 'noise' in data revisions is central to understanding the choice of policy instrument in the Poole (1970) model.
False. The 'news' vs 'noise' distinction is about the nature of data revisions over time (Aruoba, 2008). The Poole model is a static analysis of how a policymaker should choose an instrument in the face of unobservable IS or LM shocks. It does not deal with the process of data revision. (Source: Poole, 1970; Aruoba, 2008)
51. Certainty equivalence implies that a policymaker can ignore the variance of additive shocks because their policy choice does not affect it.
True. With additive uncertainty, the variance of the outcome is solely determined by the variance of the external shock(s), which is outside the policymaker's control. Therefore, the optimal strategy is to focus only on what can be controlled: the expected value of the outcome. (Source: EC3115 Subject Guide, 12.8.1)
52. If money demand becomes highly unstable due to financial innovation, Poole's analysis would recommend a shift towards targeting the money supply.
False. High instability in money demand corresponds to large LM shocks. In this case, Poole's analysis recommends targeting the interest rate to insulate the real economy from this financial volatility. (Source: Unit L Lectures, Slide 15)
53. The 'policy constraint' in the Brainard model is linear because the mean and standard deviation of output are both linear functions of the policy instrument.
True. Given \(\bar{y} = \bar{g}X\) and \(\sigma_y = \sigma_g X\), we can write \(\bar{y} = (\bar{g}/\sigma_g)\sigma_y\). This shows a linear relationship between the expected value (mean) of output and its standard deviation, which traces out the linear policy constraint. (Source: Unit L Lectures, Slides 30-31)
54. Aruoba (2008) finds that data revisions for real output growth in the US are best characterized as 'news'.
False. While Mankiw and Shapiro (1986) had earlier concluded that GNP revisions were 'news', Aruoba (2008), using a longer and different dataset, finds that for annual real output growth, both the news and noise hypotheses are rejected, leading to an ambiguous conclusion. (Source: Aruoba, 2008, p. 328)
55. When a central bank targets the money supply, an unexpected increase in government spending (a positive IS shock) will lead to a higher interest rate and higher output.
True. The positive IS shock shifts the IS curve to the right. With a fixed money supply, the LM curve is stationary. The new equilibrium occurs at the intersection of the new IS curve and the old LM curve, which is at a higher level of both output and the interest rate. (Source: Unit L Lectures, Slide 5)
56. The optimal policy under parameter uncertainty never involves fully closing the output gap, unless the policymaker is certain about the policy's effect.
True. The optimal policy is attenuated by the factor \(\bar{g}^2 / (\sigma_g^2 + \bar{g}^2)\). This factor is always less than 1 unless the parameter uncertainty \(\sigma_g^2\) is zero. Therefore, the expected gap is never fully closed. (Source: EC3115 Subject Guide, eq. 12.10)
57. The 'Great Moderation' period of lower macroeconomic volatility coincided with data revisions becoming smaller and more 'well-behaved'.
False. Aruoba (2008) finds the opposite. In the post-1984 period (coinciding with the Great Moderation), the 'noise-to-signal' ratio for many variables was higher, and the degree of predictability was greater, suggesting revisions became *less* well-behaved. (Source: Aruoba, 2008, p. 336)
58. In the Brainard model, the policymaker's indifference curves in (mean, standard deviation) space are concentric circles around the target output level.
True. A quadratic loss function \(L = (y - y^*)^2\) implies an expected loss of \(E[L] = (\bar{y} - y^*)^2 + \sigma_y^2\). An indifference curve is a constant level of loss, so the equation \((\bar{y} - y^*)^2 + \sigma_y^2 = \text{constant}\) defines a circle in the (\(\sigma_y, \bar{y}\)) plane centered at \((\sigma_y=0, \bar{y}=y^*)\). (Source: Brainard, 1967, p. 417)
59. If the economy is prone to IS shocks, targeting the interest rate exacerbates output fluctuations because it requires the central bank to accommodate the shocks fully.
True. With an interest rate target, a positive IS shock that pushes output up would also tend to raise the market interest rate. To maintain its target, the CB must increase the money supply, shifting the LM curve right. This accommodates the IS shock, leading to a larger increase in output than would have occurred under money supply targeting. (Source: Unit L Lectures, Slide 7)
60. The existence of long and variable lags in monetary policy is a form of additive uncertainty.
False. Long and variable lags are a primary example of *parameter uncertainty*. The 'variable' nature of the lag means the timing and magnitude of the policy's effect (the multiplier over time) are not known with certainty. (Source: Unit L Lectures, Slide 2)
61. According to Aruoba (2008), benchmark revisions of data are a key reason why revisions are 'well-behaved'.
False. Aruoba (2008) notes that benchmark revisions, which involve methodological changes, can distort the historical picture and are a reason why simply using the latest available data as the 'final' value might be problematic. He focuses on revisions outside of major benchmark changes. (Source: Aruoba, 2008, p. 323)
62. In the Poole model, if the IS curve is vertical (interest-inelastic), the choice between targeting the money supply and the interest rate does not matter for output stabilization.
True. If the IS curve is vertical, output is fixed at a level independent of the interest rate. Therefore, shocks to the LM curve that change the interest rate will have no effect on output. The choice of instrument is irrelevant because monetary policy, which works through the interest rate channel, cannot affect output at all. (Source: Poole, 1970)
63. Brainard's (1967) analysis implies that as econometric models improve and the standard errors on policy multipliers decrease, central banks should act more aggressively.
True. A decrease in the standard error of a multiplier corresponds to a decrease in parameter uncertainty (a smaller \(\sigma_g^2\)). According to Brainard's formula, a smaller \(\sigma_g^2\) means the optimal policy becomes less attenuated and more aggressive, moving closer to the certainty-equivalent policy. (Source: EC3115 Subject Guide, Section 12.8)
64. If a central bank has only one policy instrument but two targets (e.g., inflation and unemployment), Brainard's analysis suggests it can achieve both targets perfectly as long as the instrument's multiplier is known.
False. This relates to the Tinbergen rule. Even with certainty, one instrument can generally only achieve one target. With uncertainty, the problem is worse. The policymaker must trade off deviations from both targets, and all instruments would be used in this trade-off. (Source: Brainard, 1967, p. 421)
65. An unexpected stock market crash that reduces consumer wealth would be modeled as a negative LM shock in the Poole framework.
False. A stock market crash affects wealth and confidence, which directly impacts consumption and investment decisions at any given interest rate. This is a shock to the goods market, represented as a negative shock to the IS curve. (Source: EC3115 Subject Guide, Section 12.7)
66. The optimal policy response under parameter uncertainty depends on the policymaker's degree of risk aversion.
True. Brainard's use of a quadratic loss function implies a specific form of risk aversion. A different utility function (e.g., one that is more risk-averse) would lead to an even more cautious policy. The quadratic assumption makes the trade-off clear, but the general principle holds for any risk-averse policymaker. (Source: Brainard, 1967, p. 413)
67. Aruoba (2008) suggests that simply using the third or fourth data release, instead of the initial one, would eliminate the problems of bias and predictability.
False. Aruoba's analysis of intermediate revisions shows that the problems persist. He states, 'We can also infer from our results that simply ignoring the initial announcement and using the second or third announcement would not eliminate the problems with revisions.' (Source: Aruoba, 2008, p. 338)
68. In the Poole model, if the LM curve is horizontal ('liquidity trap'), targeting the money supply is more effective than targeting the interest rate.
False. In a liquidity trap, the LM curve is horizontal, meaning monetary policy is ineffective; changes in the money supply have no effect on the interest rate. In this region, both policies are equally ineffective at influencing output through the interest rate channel. (Source: Poole, 1970)
69. The diversification benefit of using multiple policy instruments in Brainard's model is greatest when the uncertainty in their effects is perfectly positively correlated.
False. The diversification benefit is greatest when the uncertainty in their effects is negatively correlated. A perfect positive correlation means the instruments' impacts go up and down together, offering no hedge. A negative correlation means when one is unexpectedly strong, the other is unexpectedly weak, stabilizing the overall impact. (Source: Brainard, 1967, p. 421)
70. The certainty equivalence principle is a behavioral assumption about policymakers.
False. It is not a behavioral assumption but a normative principle derived from optimal control theory under specific conditions (quadratic loss, linear model, additive uncertainty). It describes how a rational policymaker *should* behave under those conditions. (Source: EC3115 Subject Guide, Section 12.3)
71. If the final revision \(r_t^f\) is defined as \(y_t^f - y_t^{t+1}\), the 'news' hypothesis can be tested by checking if the coefficient \(\beta_2\) is equal to 1 in the regression \(y_t^f = \alpha_2 + \beta_2 y_t^{t+1} + v_t^2\).
True. If the initial announcement \(y_t^{t+1}\) is a rational forecast of \(y_t^f\), then in a regression of the final value on the initial value, the constant \(\alpha_2\) should be 0 and the slope \(\beta_2\) should be 1. This is the test for 'news' used by Mankiw and Shapiro (1986) and discussed in Aruoba (2008). (Source: Aruoba, 2008, p. 327)
72. Brainard's model suggests that even if an instrument is known to have a very weak effect on average (low \(\bar{g}\)), it should still be used if its effect is known with high certainty (low \(\sigma_g^2\)).
True. The choice of instruments is about finding the optimal portfolio. A weak but reliable instrument can be very valuable for hedging the uncertainty of a stronger but more volatile instrument. Its reliability can make it an important part of the optimal policy mix. (Source: Brainard, 1967)
73. A central bank that smooths interest rate changes, as discussed by Sack (2000), is behaving consistently with the principle of certainty equivalence.
False. Interest rate smoothing (making small, sequential adjustments) is a form of gradualism. This behavior is a direct contradiction of certainty equivalence, which would call for immediate and full adjustments. Smoothing is, however, consistent with optimal behavior under parameter uncertainty, as analyzed by Brainard. (Source: EC3115 Subject Guide, Section 12.9, 12.8.2)
74. In the Poole model, if the IS and LM shocks are of roughly equal variance, the choice of instrument has little consequence for output stability.
False. The optimal choice depends not just on the variances (\(\sigma_\epsilon^2, \sigma_\eta^2\)) but also on the structural parameters (b, d, e). The criterion for preferring one instrument over the other is whether \(\sigma_\epsilon^2\) is greater or less than \((\frac{b}{d})^2 \sigma_\eta^2\). Even with equal variances, the choice will matter if the structural parameters are not equal in a specific way. (Source: Poole, 1970)
75. The finding that data revisions are predictable means that private agents can, in principle, form better forecasts of the true state of the economy than the government's initial announcements.
True. If revisions are predictable using publicly available information, it means the initial announcement did not efficiently incorporate that information. A rational agent could use that same information to augment the initial announcement and create a statistically superior forecast of the final, true value. (Source: Aruoba, 2008)
76. Parameter uncertainty can be completely eliminated by using better econometric techniques.
False. While better techniques can reduce parameter uncertainty (i.e., lower the standard errors of estimates), it can never be completely eliminated. All models are simplifications of reality (model misspecification), and data is finite and measured with error. Some level of uncertainty about the true structure of the economy will always remain. (Source: EC3115 Subject Guide, Section 12.8)
77. When a central bank targets the interest rate, an unexpected surge in money demand (positive LM shock) will lead to a contraction in the money supply.
False. To keep the interest rate from rising in the face of higher money demand, the central bank must *increase* the money supply to satisfy that extra demand. (Source: Unit L Lectures, Slide 13)
78. The core insight from Brainard (1967) is that the correlation between a policy's impact and exogenous shocks is more important than the variance of the policy's impact.
False. The core insight is about the variance of the policy's impact (the multiplier uncertainty). The optimal policy is P* = g / (1 + V^2), which shows the importance of the coefficient of variation (V). The correlation with other shocks is a secondary, though interesting, complication, not the core insight. (Source: Brainard, 1967, p. 415)
79. Aruoba (2008) finds that revisions to inflation are less predictable than revisions to real output.
True. Comparing the results in Aruoba's Table 2 (Panel A), the R-squared values for the forecasting regressions are generally higher for real output measures than for inflation measures, indicating that revisions to real output are more predictable from past data. For example, the R-squared for annual real output is 0.19, while for annual inflation it is 0.05. (Source: Aruoba, 2008, Table 2, p. 331)
80. Poole's analysis is best described as a guide for the strategic, long-run design of monetary policy frameworks.
False. Poole's analysis is about the choice of a short-run, tactical operating target (interest rate vs. money supply) to stabilize the economy against high-frequency shocks. It is not about the long-run strategic design of policy, such as choosing an inflation target. (Source: Poole, 1970)
81. If a policy instrument's effect (g) is uncertain and negatively correlated with an additive shock (u), it makes that instrument more attractive for stabilization.
True. A negative correlation means that when the additive shock is unexpectedly expansionary (u is high), the policy's impact is likely to be unexpectedly weak (g is low). This creates a natural hedge. The instrument is 'automatically' less powerful when it's needed less and more powerful when it's needed more, which is a desirable property that reduces overall output variance. (Source: Brainard, 1967, p. 416)
82. The fact that central banks use large, complex macroeconometric models is evidence that they do not face parameter uncertainty.
False. The very existence of standard errors, confidence intervals, and competing models within central banks is direct evidence *of* parameter uncertainty. They know their models are imperfect approximations and that the estimated coefficients are not the true values. (Source: EC3115 Subject Guide, Section 12.1)
83. In the Poole model, if the money demand function is highly interest-elastic (LM curve is flat), LM shocks will have a large impact on output under a money supply target.
False. If the LM curve is flat, a given shift in the LM curve (due to an LM shock) will translate into a small change in the interest rate for any given level of Y. This, in turn, will have a small effect on output. The impact of LM shocks is larger when the LM curve is steeper. (Source: Poole, 1970)
84. The optimal degree of policy caution in Brainard's model is independent of the size of the output gap.
True. The degree of caution, or the attenuation factor \(\bar{g}^2 / (\sigma_g^2 + \bar{g}^2)\), depends only on the mean and variance of the multiplier (i.e., its coefficient of variation). It does not depend on the size of the gap \(y^*\). The *level* of the policy instrument depends on the gap, but the *fraction* of the gap to be closed does not. (Source: EC3115 Subject Guide, eq. 12.10)
85. Aruoba's (2008) finding of a positive mean for data revisions suggests that initial estimates of GDP growth tend to be revised downwards over time.
False. A positive mean revision (where revision = final - initial) means that the final value tends to be higher than the initial value. This implies that initial estimates of GDP growth are, on average, revised *upwards*. (Source: Aruoba, 2008, p. 324)
86. If the IS curve is horizontal, interest rate targeting is the only effective policy.
False. If the IS curve is horizontal, output is infinitely sensitive to the interest rate. In this extreme case, money supply targeting would be extremely powerful (a tiny change in M would cause a huge change in Y), while interest rate targeting would be impossible to implement as it would lead to an indeterminate level of output. (Source: Poole, 1970)
87. Brainard's (1967) work implies that policymakers should discard instruments with highly uncertain impacts.
False. Brainard's work shows that even highly uncertain instruments can be valuable as part of a policy portfolio, especially if their impacts are negatively correlated with other instruments or shocks. They should be used cautiously, not discarded. (Source: Brainard, 1967, p. 421)
88. The certainty equivalence principle is most applicable in a stable economic environment with a long history of reliable data.
True. In such an environment, the central bank can be more confident in its model parameters (low parameter uncertainty), so the primary source of error is unpredictable additive shocks. In this case, the assumptions for certainty equivalence are more closely met. (Source: EC3115 Subject Guide, Section 12.8.1)
89. According to Poole, the main reason central banks switched to interest rate targeting was the increasing predictability of money demand.
False. The reason was the increasing *unpredictability* (instability) of money demand, which corresponds to larger and more frequent LM shocks. This made money supply targeting volatile and interest rate targeting the superior choice for output stabilization. (Source: Unit L Lectures, Slide 15)
90. The presence of parameter uncertainty turns the policymaker's problem into one of portfolio choice.
True. As Brainard (1967) framed it, when instruments have uncertain impacts, the policymaker must choose a 'portfolio' of instruments to achieve the best combination of expected return (closing the target gap) and risk (the variance of the outcome). (Source: Brainard, 1967, p. 418, footnote 7)
91. Aruoba (2008) finds that data revisions are larger and more predictable during economic recessions.
True. Aruoba's forecasting model includes the change in the unemployment rate to capture business cycle effects. He finds it to be a significant predictor, noting that a 1% increase in unemployment would cause a downward revision to output growth of nearly 1%, suggesting revisions are systematically different during recessions. (Source: Aruoba, 2008, p. 330)
92. In the Poole model, if the goal is to stabilize interest rates, then targeting the interest rate is always the optimal policy.
True. This is tautological. If the sole objective is to have a stable interest rate, then a policy of fixing the interest rate will achieve that objective with zero variance. Poole's analysis, however, assumes the objective is to stabilize output. (Source: Poole, 1970)
93. Brainard's principle of policy attenuation is irrelevant if the policymaker's loss function is linear rather than quadratic.
False. While the quadratic function provides a specific, tractable result, the general principle holds for any risk-averse policymaker. A linear loss function implies risk neutrality, which is a very strong and unrealistic assumption. For any policymaker who dislikes volatility, parameter uncertainty will induce some degree of caution. (Source: Brainard, 1967)
94. The 'news' hypothesis of data revisions is consistent with the rational expectations hypothesis.
True. The 'news' hypothesis states that initial announcements are rational forecasts and revisions are the result of new, unpredictable information. This aligns perfectly with the rational expectations hypothesis, which posits that agents use all available information efficiently, and their forecast errors are therefore random and unpredictable. (Source: Aruoba, 2008, p. 320)
95. If IS shocks and LM shocks are negatively correlated, the case for targeting the money supply is strengthened.
True. Under money supply targeting, a positive IS shock (raising Y) and a negative LM shock (raising Y) would tend to offset each other if they occurred together. The negative correlation means the shocks act as natural stabilizers for each other, reducing the variance of output under a money supply target and making it a more attractive option. (Source: Poole, 1970)
96. The optimal policy in the Brainard model depends on the sign of the output gap.
False. The formula for the optimal instrument setting, \(X = \frac{\bar{g}y^*}{\sigma_g^2 + \bar{g}^2}\), shows that the level of X is proportional to the target \(y^*\) (the gap). However, the *strategy* or *degree of aggressiveness* (the fraction \(\bar{g}^2 / (\sigma_g^2 + \bar{g}^2)\)) is independent of the sign or size of the gap. (Source: EC3115 Subject Guide, eq. 12.10)
97. Data uncertainty (additive uncertainty) is the primary justification for gradualism in monetary policy.
False. Additive uncertainty alone leads to the certainty equivalence result, which calls for immediate, not gradual, policy action. The primary theoretical justification for gradualism is parameter (multiplicative) uncertainty, as shown by Brainard. (Source: EC3115 Subject Guide, Section 12.9)
98. Aruoba (2008) finds that revisions for government spending are a major source of the 'noise' in overall GDP revisions.
False. Aruoba (2008) breaks down real output into its components and finds that consumption of durables and exports are two components with particularly large and significant mean revisions, suggesting they are major sources of the revision properties of the aggregate. (Source: Aruoba, 2008, p. 338)
99. In the Poole model, the relative benefit of money supply targeting over interest rate targeting increases as the interest elasticity of investment (the 'b' parameter) increases.
False. The choice depends on the ratio of variances. The variance under money supply targeting is proportional to \(b^2\sigma_\eta^2\). A larger 'b' (a flatter IS curve) means that any given change in interest rates (caused by an LM shock) has a larger effect on output. This makes money supply targeting more risky in the face of LM shocks, strengthening the case for interest rate targeting. (Source: Poole, 1970; EC3115 Subject Guide, eq. 12.5)
100. The key takeaway from Brainard (1967) is that more uncertainty always leads to less active policy.
True. This is the essence of Brainard conservatism or the attenuation principle. Uncertainty about the effects of policy (parameter uncertainty) should lead rational, risk-averse policymakers to use their policy instruments more cautiously and less aggressively than they would in a world of certainty. (Source: Brainard, 1967; EC3115 Subject Guide, Section 12.8)