Economics 304 - Winter 2013
Monetary Policy in Canada and the US: Part 2
n oi tcudort1nI
In the previous lecture, we present a detailed analysis onayoiiilmtdi
Canada. While the objective of monetary policy in Canada and in many other countries is simple,
that is to target a given level of the CPI, achieving this objective is complex. In this lecture, we
give an overview of some of the data that the sta▯ and governingcouncilusetocometoadecision
regarding the level at which they should set the overnight ra te.
yci lo praten mnatniatre cU
In formulating monetary policy, central banks have to deal with substantial uncertainty. Uncer-
tainty is often pervasive and can come from numerous sourcesand can take many forms. Uncertainty
can be grouped under three categories: data, model and shheaudycoso-
tain errors and are often subject to revisions. For example,GDP in Canada published by Statistics
Canada every quarter is often revised as new information is revealed, new methodologies are im-
plemented and/or mistakes are discovered. As central banks have to take decisions in real time
but use data that are often revised, this source of uncertain ty complicates the conduct of monetary
Central banks have to make forecasts about the economy and they use various models to forecast
important macro variables such as inﬂation, GDP and the exchange rate. The forecasts will thus
heavily depend on the types and accuracy of the models they us e. As models are not perfect,
there is always an inherent uncertainty when using them. There is also uncertainty regarding the
relationship between variables in a given model and the valu es that these variables take.
Parameter uncertainty arises when central banks are unsrow changes in one variable a▯ect
the other (size and direction). For example, how much does inﬂation and GDP change following
a 1% increase in the interest rate? Di▯erent models and di▯erent methodologies will give di▯er-
ent answers. In addition, often the conﬁdence interval surrounding point estimates is large and
the relationship between variables can change over time due to structural changes. This type of
uncertainty also complicates the conduct of monetary polic y.
Typically central banks respond more cautiously the bigtre uncertainty regarding the pa-
rameters of the model. This is the essence of “Brainard princ iple”. The Brainard Principle basically
states that policy-makers should exhibit conservatism and be more cautious in the face of parameter
Central banks also face uncertainty regarding the nature ainfnmik.
They observe changes in macro variables but often do not know the cause of the changes. As
the policy response of a central bank can depend on the nature and duration of the shock, it is
therefore important for them to know the nature of the shock. For example, suppose the central
bank observes a signiﬁcant increase in real wages. If real wa ges are increasing because of a shock
1 to nominal wages but production remains the same, then this increase in wages will likely trigger
inﬂation in the future as ﬁrms will pass on the rising costs to consumers. In this case, the central
bank should be forward looking and take the appropriate acti ons to contain any future increases
in prices. On the other hand, if real wages are increasing because of a rise in productivity, in this
case, the central bank should not react as this increase in wages is not likely to trigger inﬂation in
Central banks deal with these various sources of uncertaii aumrofa. Mey
policy nowadays is conducted in a era of rich data. Central banks take advantage of this by
analyzing data on various sectors and parts of the economy. The sta▯ at the Bank of Canada
monitors a large number of data. While the focus is on macro data, they also pay a lot of attention
to micro data on households and ﬁrms.
Central banks also use di▯erent models to forecast the economy. Since there is a great deal
of uncertainty regarding the forecast of any given model, central banks typically employ di▯erent
types of models (size, transmission mechanism). The sta▯ tah te Bank of Canada uses various
models to provide information to governing council. Most of the models used at the Bank of
Canada are structural models. The main projection and policy-analysis model used at the Bank
of Canada is called TOTEM (Terms-of-Trade Economic Model). It is an open-economy, dynamic
stochastic general equilibrium model (DSGE) that is estimated to provide forecasts and policy
recommendation to governing council.
They pay particular attention to the various measures of pressures on capacity such as the
output-gap, capacity utilization, vacancy rates and unit labour cost. The output-gap is a very
important variable central banks pay attention to as the latter is assumed to be a key driver of
yci lo praten ofoataehT
In this section, we present some of the information and data used by the Bank of Canada in
the formulation of monetary policy. The approach followed beBakofCndagdighe
setting of the target overnight interest rate is based on models and data pertaining to the Canadian
and overseas economies. One of the central components of thiosiamarbigbyhe
Bank sta▯ to Governing council. This brieﬁng is focused around the analysis prepared by the sta▯
based on forecasts and projections coming from the di▯erent models used by the Bank, in particular
TOTEM, information from surveys conducted by the regional o▯ces (Senior Loans O▯cer Survey
and the Business Outlook Survey), information from ﬁnanciala,i ilrmsd
households credit data and information from ﬁnancial marke ts, in particular market expectations
regarding inﬂation and interest rates.
We present an overview of the di▯erent types of information that the Bank of Canada use in
coming to a decision regarding interest rates. We begin with measures of capacity pressures and
focus on the output-gap which is one of the key variable they use.
3.1 The output-gap
This section of the notes borrows partly from “Information and Analysis for Monetary Policy: Coming to a
Decision” by Ti▯ Macklem, Bank of Canada Review Summer 2002.
2 The output-gap is the di▯erence between the economy’s actual output (GDP) and the level of
production it can potentially produced. This potential out put represents the amount of output
an economy can achieve with existing labour, capital, and technology without putting sustained
upward pressure on inﬂation.
The gap is positive when actual output exceeds the economy’s potential and negative when
actual output is below potential output. A positive output gap is also referred to as excess demand
and a negative output gap is referred to as excess supply. Since positive (negative) output-gaps
are associated with excess demand (supply), they are often used as indicators of capacity pressure
in the economy. Excess demand and hence a positive output-gaps ifnadasnel
indication of increases in future inﬂation.
When spending in the economy is high in relation to capacity,this tends to put upward pressure
on prices. Conversely, a low rate of spending tends to put downward pressure on prices. This
relationship can also be expressed in the reverse manner—if the rate of inﬂation begins to increase,
it is typically a sign that spending levels are approaching the economy’s level of potential output
and that output growth is not sustainable. Conversely, if therateofinﬂationtlywelosten
expectations, this is a sign of continuing slack (excess capacity) in the economy.
The Bank of Canada is concerned about both too much and too little demand in the economy
when either puts sustained upward or downward pressure on prices (remember the Bank of Canada
tries to keep inﬂation within the 1-3% and, aiming for the middle of the band). Thus, when the
output gap is thought to be small and demand is seen to be increasing faster than potential output,
the Bank of Canada will typically act to tighten monetary conditions to curb demand and the
inﬂationary pressures. If the economy can be kept from overheating, then it will be less likely that
even tighter monetary conditions will be required later to control inﬂation.
3.2 The output-gap and inﬂation
We saw that a positive output-gap is an indication of eex mcensd in the economy and as we
mentioned earlier, a positive output-gap is often a signal that there will be pressure on prices to
increase in the future. The link between the output-gap and inﬂation can be summarized by the
Phillips curve. We will develop the Phillips curve in the coming lectures. For now, assume that
the output-gap at time t is given by x ,niainy ▯ .Teirttenk
of Canada use in their models take the following form:
▯t= ▯E ▯t t+1 + ▯x t ▯ t (1)
▯ and ▯ are parameters and ▯ it an exogenous shock to inﬂation. The above equation is a forward-
looking Phillips curve. In this equation, inﬂation at time t depends on expected future inﬂation,
the output-gap and an inﬂation shock. If x is positive, the output-gap is positive, implying there
is an excess demand in the economy. If this is the case, there will be pressure on inﬂation to
increase. Inﬂation also depends on expected inﬂation, a key driver. The Phillips curve is a very
good representation of how excess demand or supply and inﬂation expectations a▯ect inﬂation.
This forward-looking Phillips curve implies that the central bank can e▯ectively reduce inﬂation
through two channels. By reducing inﬂation expectations and/or by increasing the policy rate so
that the output-gap becomes negative. A reduction in inﬂation expectations can come from the
policy actions of the central bank but it can also come from an nouncements from the central bank.
3 The latter is possible if and only if the central bank is very credible. Credibility as we will see later
on is a key driver of inﬂation expectations.
3.2.1 Measuring the output gap
The output-gap is not easy to measure and there is a great deal of uncertainty around estimates of
the output-gap. This is because potential output is an unobserved variable that depends on numer-
ous factors (productivity, technology, population growth,etc...)Bothelofptialoutput
and the output gap are estimates, and therefore there is majoruncertyintheircalorlationF
this reason, the Bank of Canada also uses other indicators of capacity when formulating policy.
Figure 1 shows estimates of the output-gap for Canada from 2007-2012Q4 as well as two mea-
sures of capacity from the Business Survey Outlook that the Bank of Canada conducts every
quarter. The red line is an estimate of the output-gap by the Bank of Canada, the blue bars
indicate the percentage of participants in the Business Outlook survey that have di▯culty meeting
demand/sales and ﬁnally the green bars is percentage of ﬁrms in the survey indicating experiencing
The output-gap became very negative (about 3%) at the height of the ﬁnancial crisis in 2009,
indicating that there were pressures on inﬂation to fall below the target during this period. The
output-gap is still negative according to the estimates of the Bank, indicating that there is still
some slack or excess supply in the economy. As a result, this indicator implies that there is some
mild pressure on prices to fall below the target of 2%on by the Bank of Canada.
3.2.2 Other measures of capacity
On top of the output-gap as a measure of capacity and degree of excess demand and supply, the
Bank of Canada uses many other indicators of capacity. Most oft hma epbedbySii
Canada. These include unit labour costs, aggregate stock-tosa est,maufungspmes,
labour shortage from the regional survey and various measures of cost and inﬂation.
The blue and green bars in Figure 1 shows two such measures. They are the response from
participants in a survey that the Bank conducts every quarte ra mngbuisi Cnd. hse
two measures respectively provide some indication whethﬁ erms are having problems meeting
demand and whether they are experiencing labour shortages. If ﬁrms are unable to meet demand,
this would indicate that there will be pressure on prices to increase as this is an indication of excess
demand. In the same spirit, if ﬁrms are having di▯culties hir ing workers, this shortage will put
pressure on wages to rise and hence on inﬂation if higher wage sfediopi.
Figure 2 present di▯erent measures of labour cost that the Banko fCndamnsey
quarter. These di▯erent measures of unit labour cost gives an indication of wage pressures and
hence wage inﬂation in Canada. If wages are growing faster th an productivity, this may be a sign
of wage inﬂation in the economy.
In addition total or headline CPI, the Bank of Canada monitorsv sormasof
inﬂation to have an indication if there are pressures on prices to increase, Figure 3 shows various
measures of inﬂation. As we mentioned in the previous lecture, core inﬂation is the measure of
inﬂation that the Bank of Canada follows closely and uses for forecasting and policy-analysis.
However, the Bank also follows several other measures of inﬂation such as CPIXFET and CPIW.
4 CPIXFET is CPI excluding food and energy and indirect taxes, whereas CPIW adjusts each CPI