Aid
ES10002 (at least that is the number this month!!) Introductory Macroeconomics
Lecturer John Hudson, email j.r.hudson@bath.ac.uk..
Organisation 2
x 1 hour lectures per week
10
x 1 classes per semester (ALL students)
Assessment 60%
end of semester examination
20%
for each of two essays for class tutors
Reading R.A.Lipsey and A. Chrystal Principals of Economics
(ninth edition)
Griffiths
and S. Wall (eds) Applied
Economics
Begg Fischer and Dornbusch
A
series of handouts in the Short Loan section of the Library
ALSO SEE EXCEL SPEADSHEET HANDOUT: http://staff.bath.ac.uk/hssjrh/macrodats.xls
& TALK TO
NOTE THESE NOTES WILL BE CONTINUALLY EVOLVING ONLY PRINT A FEW
LECTURES AT A TIME.
Web Sites:
HM Treasury in
General economics web site: http://dspace.dial.pipex.com/town/parade/rcb48/famous.htm
Economist Website: http://www.economist.com/
American Economics Association (with lots of resources) http://www.vanderbilt.edu/AEA/
Bank of
And for video clips http://www.bankofengland.co.uk/education/videos/index.htm
Several video clips relevant here, including the ones on inflation and also the one on what the Bank of England does
Also of passing interest: http://research.stlouisfed.org/publications/review/06/07/Kotlikoff.pdf
For recent data on the whole of the
The New Palgrave Dictionary of Economics, described as the world's most extensive and authoritative economics reference resource. The New Palgrave Dictionary of Economics, 2nd edition, edited by Steven N. Durlauf and Lawrence E. Blume, contains over 1,850 articles by more than 1,500 of the world's leading economists. http://www.dictionaryofeconomics.com/dictionary
See also these data sources: http://www.bath.ac.uk/library/info/ec.html
Also see IFS later
Additional
Material Specific use of Winecon
will not be done in the lectures or classes. But students can make appropriate
use of it on their own initiative. Students will, however, make constant use of
the IFS’s Virtual Economy. Amongst other things this
includes a macroeconomic model of the
References:
L+C: Lipsey and Chrystal
VE Virtual Economy on the Institute for
Fiscal Studies Website: http://www.ifs.org.uk/
And choose virtual economy option on the left, alternatively:
http://www.bized.co.uk/virtual/economy/
HMT: HM Treasury in
GE: General economics web site:
http://dspace.dial.pipex.com/town/parade/rcb48/famous.htm
See also:
Essay on US Trade deficit and trade deficits
in general: http://www.bepress.com/cgi/viewcontent.cgi?article=1020&context=ev
Course Outline
- Introduction to Macroeconomics L+C Ch. 20
- Circular Flow of Income and Expenditure L+C Ch.20
- National Income accounts RP
- The Consumption Function and the multiplier L+C Ch. 21
- For above: http://ingrimayne.com/econ/Keynes/Overview12ma.html
- The Investment Function and the accelerator L+C Ch. 21
- The money supply and money institutions L+C Ch. 25 and Ch 26 http://en.wikipedia.org/wiki/Money_supply
- Output determination in a closed economy, aggregate expenditure,
the IS-LM model
- Exports and imports and the balance of payments L+C Ch 28
- Exchange rates, purchasing power parity, The Purchasing Power
Parity debate, Alan Taylor & Mark Taylor, Journal of Economic
Perspectives, Vol. 18, Fall 2004
- Unemployment Ch. 3, Hudson (1)
- Inflation L+C, Ch 30 Hudson(2)
- The Keynesian approach to policy making and the life of John
Maynard Keynes, VE, RP, GE
- The monetarist approach to policy making and the life of Milton
Friedman, VE, RP, GE
- The balance of payments and exchange rates L+C Ch. 28
- Macroeconomic Modelling,
- Economic growth and economic cycles: The Kondratief
and the life of Joseph Schumpeter. L+C Ch. 27 RP, GE
- The impact of bankruptcies on the macroeconomy.
Chapter 11 of the
- The CREDIT CRUNCH
- The Impact of Aid on Growth
and Poverty – Papers in Economic Journal Symposium, June 2004,
particularly Collier and Dollar, see too: http://video.web-manager.co.uk/dfid/video/wm.asp
- The Determinants of Productivity – Notes to be added to course.
NOTE: Two links for TV recordings of
lectures. These will not be given ‘live’. BUT do form part of the course and
may be examined on as with any other topic. The first is on modelling the
second unemployment.
http://wms.bath.ac.uk/live/Economics/Modelling.wmv
http://wms.bath.ac.uk/live/Economics/unemployment.wmv
http://staff.bath.ac.uk/hssjrh/FISCAL
POLICY.ppt
For a video on Keynes see:
or
ESSAY Number ONE: Deadline: Friday 11 March
MIDDAY (i.e. 12
Compare what has been happening to growth in the
Word Limit 1500 Words; people say what does this mean, is 1600 words OK? NO. 1500 words, is the limit, 1501 words is over the limit. This excludes the title, figures/diagrams, tables and references. But includes footnotes. What is recent? That is up to you to decide. But a silly choice will lose marks
With this first essay (but not the second)
you will get back a mark sheet relating to the following:
·
Essay style,
including appropriate use of paragraphs, sentences (not too short and not too
long and not continually beginning with the same word(s)), spelling, etc. A
good structure should contain an introductory paragraph outlining what you will
do and defining anything that needs to be defined. In subsequent paragraphs you
should deal with a point or a point of view, finish and then move onto the
next. (20%)
·
Your grasp of the
economics. (40%)
·
Originality and
breadth of the literature referenced. You should cover all the relevant
material AND perhaps give references that were not given to you. But beware not
all journals are of equal stature nor are all articles. Also too look at the
IMF, World Bank, OECD and Central Bank websites. (20%)
·
The ‘feel’ of an
essay. When you see a film or play you like, or a piece of music, or a book, you
know that it is impossible to totally define in words why you like it. Yes it
was exciting, well acted, well played, etc. But at the end of the day there is
something which distinguishes a good piece of work from a poor one which it is
impossible to pin down. So it is with essays. We cannot give you a cookbook of
recipes which constitute a ‘good economist’, or rather the cook book will only
take you so far. (20%)
ESSAY Number TWO: Deadline Friday
April 15, MIDDAY. (i.e. 12 noon. This topic is
dependent on the program being available at the time of the project). Back up your laptop on a flash drive, in
particular keep copies of all your essays and important work. Losing work because of laptop failure will not be acceptable
as an excuse for late work.
The student is to
log into the IFS Website (see the website at the beginning of these notes with
other web references) and run their Virtual Economy model of the economy. I
will give you the following specific changes to make. These will be different
for different students (the foruma for this is given
below).
- Change the basic rate of Direct Taxation by x1
(e.g. 22 to 26)
- Change Indirect Taxation by x2 (VAT)
- Change Government expenditure by x3 (e.g. both
government capital spending and government current spending), i.e. all of:
|
|||
|
National
Health Service |
|
||
|
Education
|
|
||
|
Law and
order |
|
||
|
Defence |
|
||
To calculate x1-x3 take your date of birth:
e.g. 29.06.83 (29 June, 1983)
For x1 subtract second number from first (2-9
=-7). If in excess of 5 in absolute value then divide by 2 and if necessary
round up) -7 is in excess so divide by 2 = -3.5 and round up to -4. [Note a
negative increase is in fact a decrease]. This is then x1. Repeat for x2 with
third and fourth digits, =0-6=-6, divide by 2=-3 no need to round up hence
x2=-3. For x3 use final two digits: 8-3=5, No need to divide by 2 hence x3=5.
If any number is zero you may raise it to 1%. The above does leave latitude for
you to make certain decisions (e.g. what exactly is direct taxation). This is
part of the project and you are to defend those decisions and the marks will in
part depend upon the assumptions you make. If you are to change interest rates
then you should divide your number by 4.
You are to report the impact of the above on
(i) inflation, (ii) unemployment and (iii) the
balance of payments. This will be the case for each of the above on their own and for all three combined. You are then to
comment, i.e. explain, these changes.
Word Limit 1500
Words. THIS MEANS 1500 words.
Classes:
SEE EXCEL SPEADSHEET HANDOUT: http://staff.bath.ac.uk/hssjrh/macrodats.xls
Attendance at classes is compulsory. failure to attend either two classes in succession or more
than two classes in total will result in your being reported to the Director of
Studies and possibly to the university authorities. Your tutor has total
flexibility to arrange the classes as he/she feels is best.
The suggested format is that the tutor should assign to 4/5 members of the
tutorial group the specific task of presenting the class topic to the rest of
the group. The main exception to this is the macro-economic modelling exercise,
where all students should bring their results. But again the tutor has total
flexibility to design the classes as they wish. The following is only a
suggested list of topics.
1. The main factors affecting the marginal
propensity to consume in the
2. Assuming purchasing power parity, take the
exchange rate of the US$ against the pound sterling in 1980 then calculate what
the exchange rate would have been, with 1970 as the base year, had purchasing
power parity prevailed. That is you should calculate: Noute
PUS (the consumer price index in the
ER* = ER(1970)*PUS/PUK
where ER is the exchange rate of £ into $, currently about
1.60, PUK is the consumer price index in the
3. Plot the Phillips curve since 1970, i.e.
the inflation rate on the vertical axis against the unemployment rate (in %
terms) on the horizontal axis. Now assume that expected inflation in one year
(e.g. 1990) equals actual inflation in the previous year (e.g. 1989) calculate
the change in the inflation rates and plot against the unemployment rate and
comment. Plot the unemployment rate in the
4. The student is to log into the IFS Website
and run their Virtual Economy model of the economy.
- Increase Direct Taxation by 2 percentage points
- reduce the rate of interest by 1 percentage point
- Reduce Government spending by 3%
SEE THE END OF THESE NOTES FOR FURTHER
GUIDANCE
Summarise the impact upon the economy and
rationalise these changes using the theory you have learnt during the course of
the lectures.
5. The final classes might well be taken up
by preparation for the exam.
Studying at University is about working hard,
not just accepting what you are taught, but reasoning it out for yourself and
deciding whether you accept it or not. As first years that is a tall order, but
as you progress through your degree this should characterise your learning more
and more. Do not misunderstand, you are not to reject everything, and when you
do you reject a theory you should be cautious.
Most times no one theory is totally right, nor indeed totally wrong. It is
shades of grey rather than black and white. Studying at a top
University demands even more.
What you get out of it depends on what you
put into it and what others around you put into it. The value of your degree
also depends upon the reputation of
But take time too to explore the region,
http://www.castlewales.com/tintern.html
Macroeconomics Lectures: First Years
These notes are a guide to the course. They
are no substitute for reading the references nor
attending the lectures, where more will be covered.
Macroeconomics Lectures: First Years
1. Introduction to macroeconomics; the circular flow of income, GDP tables, the
difference between nominal and real variables. Log on to: http://www.imf.org/external/mmedia/view.asp?eventID=1718
Download: http://www.imf.org/external/pubs/ft/fmu/eng/2010/01/index.htm
http://www.imf.org/external/pubs/ft/weo/2010/update/01/index.htm
The national income identity:
Y = C + I + G + X – M (1.1)
Y=GDP
C=Consumers’ expenditure
I=Investment
G=Government spending
X=exports
M=imports
For discussion on measuring GDP See: http://www.statistics.gov.uk/cci/nugget.asp?ID=56
Test yourself: http://highered.mcgraw-hill.com/sites/0077099478/student_view0/chapter19/self-test_questions.html
For circular flow of income: http://www.bized.co.uk/educators/he/pearson/models/circular.ppt#316,1,Slide 1 (skip the early slides and the demand and supply curves and go straight to th slide which says Consumption, injections, withdrawals and equilibrium.
2 Consumers’ expenditure. Let us
set up a simple model
C = a + bYD (2.1)
YD=personal disposable income = (1-t)Y (2.2)
Where a = autonomous consumption, b = the marginal
propensity to consume, t = the tax rate, (1-t) = the proportion of every pound
earned retained as income after taxes. The average propensity to consume equals
C/YD = (a +bYD)/YD = a/YD + b. The marginal
propensity to consume represents the change in consumption (D C), following a change in disposable income (D Y). It is D C/D Y and in the above equation
equals b. This is an example of the simple Keynesian consumption function
(after the economist John Maynard Keynes). An example of Keynesian economic
policy would be when the government wished to increase Y by increasing C. It
could do this by cutting t, increasing YD.
Alternatives to the Keynesian consumption
function are provided by Milton Friedman’s permanent income hypothesis and Ando
and Modigliani’s life cycle hypothesis. The former postulates that consumption
can be modelled such:
Ct = a + bYDpt (2.3)
Where YDP is
‘permanent income’. In practice
Friedman proxied permanent income by a weighted
moving average of past incomes:
YDpt = g 0YDt + g 1YDt-1
+ g 2YDt-2 + g 3YDt-3 +…………….. (2.4)
where g 0 > g 1 > g 2 >……………. In
practice, however we do not estimate:
Ct = a + b(g 0YDt + g 1YDt-1
+ g 2YDt-2 + g 3YDt-3 +……………..)
(2.5)
But Ct = b 0 + b 1YDt + b 2Ct-1 (2.6)
(Note: t denotes the time period,
t is current time period, t-1 previous one and so on. The use of b ’s in an
equation like (2.6) is standard in economics. Comparing (2.5) and (2.6) we can
see that a= b 0.) In equation (2.5) the bigger the
influence of the past, the bigger will be g 1, g 2,.. etc
in relation to g 0. In equation the influence of the past
is picked up by b 2. The closer this is to 1.0 (its upper
bound), the greater the influence of past consumption on present. In terms of
Keynesian economic policy, a cut in taxes which increases YDt
will have a much more muted impact on C in (2.6) than in (2.1). Friedman’s
whole career has been spent on trying, with considerable success, to prove that
Keynesian economic policy is (i) harmful and (ii)
ineffective in controlling Y.
Question: in equation (2.6) what is the short and the long run marginal
propensities to consume (mpc)?
The consumption function is one of the most
researched concepts in economics and still the research goes on. To(2.6) we might consider adding (i)
the real rate of interest (rt) and (ii)
the rate of inflation (D Pt, generally denoted by P with a dot over
it):
Ct = b 0 + b 1YDt + b 2Ct-1 + b3D Pt +
b 4rt (2.7)
The impact of the rate of interest is obvious, it represents the returns to savings. If people
respond to these returns (i.e. they are ‘incentive savers’) then a higher rate
of interest will increase savings and thus reduce consumers expenditure (b 4<0). But there are alternative possibilities. They may
be ‘target savers’, saving for a specific amount (£1000 for a holiday) an
increase in the rate of interest will allow them to reach their target more
easily, hence they need save less and consume more (b 4 >0). The impact of inflation was first suggested by
Angus Deaton - then at
3. Investment Expenditure
One can distinguish between
inventory investment (investment in inventories or stocks of
finished goods and raw materials),
residential investment (investment in housing, i.e. new housing),
and
business investment or gross fixed capital formation (includes
public sector investment, roads, hospitals, and machines).
In general we will be focusing on the latter,
unless specified otherwise. The main factors influencing I will include the real
rate of interest [r] and Expected future retained (i.e. net of
company taxation) profits. The real rate of interest is (using the Fisher
equation) equal to nominal interest rates (n) less expected inflation (D pe):
rt = nt - D pet (3.1)
In fact it is standard to denote inflation
with a dot (· ) above the p, but it is simpler to type D p. The use of the superscript to denoted expectations is standard.
Expected profits net of taxation indicate the
potential importance of company taxation in determining investment. Taxes apart profits will depend upon how well the economy is doing
and how well it is expected to do. How to model such expectations? One
possibility – similar to the permanent income hypothesis is that we model
expected profits (P e) as a weighted moving average of past
profits:
P et
= g 0P t + g 1P t-1 + g 2P t-2 + g 3P t-3 +…………….. (3.2)
Hence (again similar to when we wanted to
model lags in the consumption function) we can use the following equation:
It = b 0 + b 1P t + b 2rt + b 3It-1 (3.3)
The lags involved in investment can be very
long indeed, investment projects take a lot of planning, etc and it may be two
years or longer before a change in interest rates has its full impact on
investment. One version of the investment function which bypasses the link
between profits and GDP is the accelerator model. The naïve accelerator is:
It = g (Yt
– Yt-1) (3.4)
Where g is the capital-output ratio.
A slightly more sophisticated version is:
It = b 0 + b1(Yt – Yt-1)
+ b 2rt + b 3It-1 (3.5)
4. Exports depend upon World output (WD, non-standard
terminology) as a proxy for output in key markets and the competitiveness
(COMP, not standard terminology) of British (to be parochial) goods. The latter
is determined by the exchange rate and prices in
COMP = ER(PH/PUS) (4.1)
Sometimes this is called the real effective
exchange rate. For example ERxPH is
the price of a British car in US dollars. E.g. car = £10,000, exchange rate
$1.60 = £1, then price in US dollars of car = $16,000. If the SAME quality of
car retails for $20,000 in the
Hence the demand function for British exports
is:
ln(Xt) = b 0 + b 1ln(ERPH/PUS)t
+ b 2ln(WDt)
+ b 3ln(Xt-1) (4.2)
ln stands for log (to the base e). Why logs?? Because when
we use logs on both sides of the equation the coefficients are elasticities. Hence b 2 is the
elasticity of British exports to World demand. Or rather it is the short-run
elasticity because of the presence of b 3ln(Xt-1).
Why include this variable? Same reason as always, it proxies
lags in behaviour. Clearly, for example if we devalue the currency to
boost exports, the full impact will not be felt immediately, Most
obviously orders will have already been placed on the basis of yesterday’s
exchange rates.
5. Imports and the balance of trade
See US balance of trade figures: http://www.bea.gov/newsreleases/international/trade/2007/pdf/trad0807.pdf and
http://www.economist.com/countries/USA/profile.cfm?folder=Profile%2DEconomic%20Data
The equation for imports is likely to be
similar to that for exports and it will follow the same logic. Basically,
ln(Mt) = b 0 + b 1ln(ERPH/PUS)t + b 2ln(Yt) + b 3ln(Mt-1) (5.1)
where we have replaced WD with Y. That is the demand for
British imports will increase with GDP. Remember one interpretation of Y is as
the total amount spent in an economy. Clearly, the more that is spent, then
other things being equal, the more will be spent on imported goods. b 1 now represents the ‘price’ elasticity of British imports.
The balance of trade is the difference
between exports and imports (X-M). It is in surplus when X>M and in deficit
when X<M. A trade deficit can be financed by (one way or another)
‘borrowing’ money from abroad. This may be explicit as for example borrowing
from the IMF or on the World capital markets. Or it might be implicit when
firms borrow on the world capital markets (where they may be able to obtain
cheaper loans than in the
Nonetheless if a government wishes to improve
its trade balance, one way is to devalue the currency. The impact on the trade
balance will be represented by the J curve (see the text book).
Named after English
political economist Alfred
Marshall (1842-1924) and Romanian-born economist Abba Lerner
(1905-1982), Marshall-Lerner principle states the conditions under which a
change in a country's exchange rate will improve its balance of payments.
See The Big Mac
index: http://www.economist.com/markets/indicators/displaystory.cfm?story_id=8649005
6. The Money Supply and Monetary
Institutions and the LM Curve http://en.wikipedia.org/wiki/Money_supply and http://en.wikipedia.org/wiki/M4_money_supply
If you were to ask the typical man in the
street what he thought the money supply was then he would probably respond that
it was the amount of notes and coins in circulation. But if you were to ask him
what he thought the definition of money was he might respond that it is used to
buy things with – in economics jargon, a medium of exchange. But this presents
a problem with the earlier definition for I can buy things with my cheque book
and also with my credit card. In fact money is defined as having the following
purposes:
It acts as a medium of exchange
It serves as a store of value,
that is you need not spend on something today, but delay that
expenditure to tomorrow, or next month. You keep your ‘money’ until you need
it.
It serves as a unit of account. A pint of beer
costs £1.80. When you go to the bar and ask the price that is what they will
say. They will not say, two loaves of bread or one
tenth of an economic lecture.
In all of these aspects money makes life
easier for firms, consumers and governments. When I want a pint of beer, I do
not have to find a publican who wants economics lectures and agree with him a
rate of exchange (an example of a BARTER economy). The government pays me in
money, which I can then use in any inn in the country.
Now it should be obvious that there can be no
one measure of money and the following summarises some of the definitions we
use:
Notes and coins outside Bank of England 25,891
Bankers’
deposits at BoE
180
M0
26,071
Notes and coins with public
21,648
Non-interest bearing bank deposits 34,295
Other retail bank deposits 331,783
Building society retail shares and deposits 96,747
M2 (now referred to as retail M4) 484,473
Wholesale bank deposits + CDs 234,660
Wholesale building society deposits + Cds
6,868
M4
726,001
M3H
848,277
M3H is a new harmonised measure created to
have standard money definitions throughout the EU. It is equal to M4 plus
residents’ foreign currency deposits in
Note: CDs sterling certificate of deposits.
However, as the Wikipedia website emphasises
the statistics have changed and now the Bank of England emphasises M0 and M4.
But the others can still be calculated.
http://en.wikipedia.org/wiki/Money_supply#United_Kingdom
Chapter 25 in L+C provides an excellent
summary of all this. The important point to note is that the BoE can control
the amount of notes and coins in circulation. But has a degree more difficulty
in controlling other types of money. Thus a high street bank can lend £1,000 to
one of its customers who adds it to her current
account thus increasing M2. However the extent to which the
Banks can do this is limited by certain ‘ratios’ e.g. the amount banks can lend
in this way cannot exceed (has an upper bound) some multiple of its cash +
deposits with the BoE. hence if the BoE
increases M0 by 10% it gives greater freedom for M4 to increase. But this is
not an exact science. Also turn to Table 6.2 etc for more details on money
supply http://www.statistics.gov.uk/downloads/theme_economy/ETSupp2006.pdf and still more from the Bank of ENGLAND: http://www.bankofengland.co.uk/statistics/ms/2009/jan/bankstats_full.pdf
and for some definitions see: http://www.bankofengland.co.uk/statistics/ms/2009/jan/background.htm
The end of cash?????? SEE: http://news.bbc.co.uk/1/hi/business/7850945.stm
We now turn to why people demand money. The
following is basically Keynes’ approach as specified in the General Theory (the
book he published in 1936)
Transactions motive People
desire money to finance their expenditures or transactions varies with Y The
greater is Y the greater will be the number of transactions and the greater the
demand for money.
Precautionary motive: these provide a cushion against the uncertainty of
life and when and where you will be called upon to pay cash. varies
(with R?)
Speculative demand. People
keep money aside for ‘a good thing’. varies with R.
(the nominal rate of interest)
The argument for the above two is simple (and
also a little simplified). If people keep money back for speculative or
precautionary purposes then they do not receive the ‘interest’ they could get
if they invested it (this is less true know than in 1936 when Keynes formulated
the theory, now you can get interest on money you keep in a current account.
Arguably the real cost is the difference between the current account and the
deposit account rate of interest as in (6.2) below, lets ignore this for the
time being)
Hence the demand for money will be a log-linear
function like:
ln(MDt) = b 0 + b 1ln(Yt) + b 2Rt
+ b 3ln(MDt-1) (6.1)
In the BoE model of the economy we have the
following long-run equation:
ln(MD) = ln(Y)
+ 0.02(RD - RS) +0.78ln(NFW/Y) (6.2)
where RD-RS = deposit rate -
Turning back to (6.1). This represents the demand for money. Equilibrium
in the money market will exist when money demand equals the money supply. As we
saw earlier we can assume, with some inaccuracy perhaps, that the Central Bank
(the BoE in the
ln(MS1) = ln(MD)
= b 0 + b 1ln(Y) +
b 2R (6.3)
Rearranging to get (6.1) on the eft hand side we can see that this becomes:
ln(MS1) - b 0 - b 1ln(Y) = b 2R
or ln(MS1)/b 2 - b 0/b 2 - b 1/ b 2ln(Y) = R (6.4)
This says that given a fixed money supply and
given a specific value for Y there will be a unique value of R which secures
equilibrium in the money markets. This relationship in (6.4) which specifies
the combinations of R and Y which secure equilibrium in the money markets is
known as the LM curve. Intuitively we have
Key paragraph; I know the above is not easy,
if you cannot understand the above algebra, you must understand the following
argument. Y1 ® TD1 ; R1
® SD1 If we have equilibrium in the money
market then TD1 + SD1 = MS1. Right now what
happens if Y increases to Y2? Y2
® TD2 (where TD2
> TD1). Hence with a fixed money supply in order to retain
equilibrium in the money market SD will have to fall which will require a
higher rate of interest. Equilibrium in the money market means that with a higher
level of GDP we must have a higher level of R. We will derive the LM curve
diagrammatically. But note the following. The LM curve slopes upwards. An
increase in the money supply means that for a given level of interest rate and
hence a fixed speculative demand, there will be more for transactions purposes
and the new equilibrium will require an increase in Y. Diagrammatically the LM
curve shifts to the right with an increase in the money supply. Note: TD,
transactions demand, SD speculative demand. We have assumed precautionary
demand does not vary with R. R= rate of interest (nominal) and Y=GDP.
7. The IS Curve
We have covered equilibrium in the money
markets. We now turn to equilibrium in the real side of the market - the goods
market. We have stressed before that we can measure GDP by counting up how much
people spend or by counting up their income, the two are the same. Total income
= Total expenditure = Total output. Let us concentrate on total income in a
closed economy (no imports or exports). People do three things with their
incomes: they spend it, they pay direct taxes with it or they save it:
Y = C + S + T (7.1)
But we have already seen, and indeed it has
been the basis of our analysis, that on the expenditure side, there are in a
closed economy three sources of expenditure:
Y = C + I + G (7.2)
Now lets simplify
things still further. Let us assume a balanced budget (T=G). It does not take a
genius to see that all of this implies that:
I +G = S +T (7.3)
We have done nothing startling so far and
indeed (7.3) can hardly be a condition for equilibrium. It is a truism, it is
true by definition. But this alone should give you cause to pause a minute and
think. Why should it be that investment which is largely done by firms should
equal savings which is primarily done by households? Why should two decisions
made by two different sets of people be equal always and all the time? The
answer is of course is that we fiddle the figures. Any mars bars left on the
shelf at the end of the year which the shop keeper had expected to sell are
part of his stocks or inventories and counted as investment. Not an investment
she he/she expected to make, not a planned investment but an investment
none the less. The same is true for unsold cars, cranes, aeroplanes, etc, etc.
The firm who makes steel pipes and sells less than expected engages in unplanned
investment. Now it should be obvious that a plausible definition of
equilibrium is when everyone’s plans are realised. In this case planned
investment equals actual investment (Ip =
I) and planned saving equals actual savings (Sp = S) (I have not
followed the argument through but it is pretty obvious that in certain cases
consumers will not be able to buy what they will have planned in which case
there will be unplanned saving). Now when everyone’s plans are realised and
(7.3) holds, as it always does, we have the following definition for
equilibrium in ‘the goods market’ or ‘the real economy’.
Ip + G= Sp +T (7.4)
We have already analysed what people plan to
save, it is the reverse of what they plan to consume. In the simple keynesian function it is a
function of disposable income:
Sp = a* + (1-b)YD
(7.5)
Sp = a* + (1-b)(1-t)Y
(7.6)
I will not do it here but this derives
straight from (2.1) and (2.2) with a* = -a. The key point and it is a pretty
obvious one is that planned saving is an increasing function of Y. Nor will
I go into this but it Is pretty obvious:
T = tY (7.7)
Now lets turn to planned investment, our
previous analysis suggested that this would be an
decreasing function of the rate of interest (R). We are simplifying here. I am
not distinguishing between the real and the nominal rate of interest and I am
not considering all the other things that may impact on investment.
IP = c + dR (7.8)
remember that because an increase in R causes a fall in IP
d will be negative. We will assume G
is exogenous, i.e its value is
fixed and is therefore outside the scope of the analysis. Right, now
using (7.4), (7.6), (7.7) and (7.8), we can see that
equilibrium in the goods market requires:
c + dR +G = a* +
(1-b)(1-t)Y + tY
Rearranging we have:
dR = a* + (1-b)(1-t)Y - c + tY
- G
R = a*/d + (1-b)(1-t)/dY - c/d + t/dY - G/d (7.8)
Now (1-b)(1-t)/d
< 0 (because d<0). Hence an increase in the rate of interest will need a
lower level of Y to retain equilibrium. Again I am aware that the above
argument is a bit convoluted, so lets give it an
intuitive spin:
Key paragraph: An increase in interest rates
will cause a reduction in planned investment. To maintain equilibrium we must
have a reduction in the total of planned savings and taxation. This requires a
reduction in Y. The IS curve associates higher interest rates with lower Y.
R ® ¯ IP
Hence to stay in equilibrium we need ¯ SP which
will only come about of Y¯
Just one final note before we do this diagramtically. According to (7.8) what impact will an
increase in G have on R? What further impact will this have on I?
We now turn to a diagrammatic derivation and
analysis of IS-LM curves.
For
IS & LM Curves
click on following links:
http://staff.bath.ac.uk/hssjrh/MACFIG1.pdf
http://staff.bath.ac.uk/hssjrh/MACFIG2.pdf
The IS curve shows the locus of points which
generate equilibrium in the goods market. Where equilibrium is defined in the
case of a closed economy (no exports or imports) as when planned investment +
government spending equates to planned savings + direct taxation. (direct taxation = income tax in the main). This is derived
algebraically above. Government spending is fixed. An interest rate of R1
results in a level of planned investment such that I + G = A (note I omit the
superscript p from Ip, denoting planned
investment, for simplicity). The 45 degree line transfers this distance to the
savings + taxation axis. (denoted A again). The level
of Y which will generate savings + taxation equal to this is Y1.
These two points Y1 and R1 lie on the IS curve. Y1
generates savings + taxation which equals the level of
I+G resulting from R1. This exercise is repeated with a lower rate
of interest such that two new points are generated on the IS curve (R2,
Y2). Join these two points together and we have the IS curve.
The student should now answer the questions:
(i) what happens if exogenous government spending increases.
Remember G is determined outside the scope of the model. An increase in G
results in an upward and parallel shift in the I + G
curve. Plot the new IS curve and you should find it shifts outwards to the
right.
(ii) What happens if the tax rate increases?
Well obviously T (total taxation increases). But S?
Well people will be left with less disposable income and hence they will save
less. An increase in the tax rate increases T and reduces S. Which effect
predominates? An example will illustrate. Let us suppose we have S = a* + (1-b)YD = a* + (1-b)(1-t)Y. Keep things simple set a* = 0.
b=0.7, t = 0.4 and Y=£1000.
S = 0.3*0.6*1000 = £180. T = tY = 0.4*1000 = £400; S+T=£580.
Now increase t to 0.5
S = 0.3*0.5*1000 = £150. T = tY = 0.5*1000 = £500; S+T=£650.
The increase in t (the tax rate) has
increased taxes by more than it has cut savings. If you think about it for a
minute you may perceive its obvious. If not and if you
are bewildered by the difficult maths just learn this an
increase in t increases S+T. This will swivel the S+T curve downwards. Again
draw the new IS curve you will see it has swivelled inwards.
What is the minimum you must know? The IS
curve is the combinations of R and Y which results in equality between planned
I + G and planned S +T. An increase in G shifts the IS curve outwards. An
increase in t swivels it inwards.
The LM curve shows the locus of points which
generate equilibrium in the money market. Where equilibrium is defined as when
the demand for money equals the supply of money. For simplicity we restrict the
demand for money to just two types: transactions demand (depends just on Y) and
speculative demand (depends just on R). This is defined diagramatically
above. Quadrant IV shows how the speculative demand for money varies with the
rate of interest. A rate of interest R1 will generate a speculative
demand SD1. The 45 degree line transfers this amount (0-SD1)
to the gap between the money supply and TD1). If the money supply is
fixed where it is, a speculative demand of SD1 will leave TD1
for transactions purposes. What level of Y will generate such a transactions
demand? Y1. Hence these two points Y1 and R1
lie on the LM curve.Y1 generates a transactions demand TD1
and R1 generates a speculative demand such that these just equal the
available money supply. This exercise is repeated for two other interest rates.
Join all 3 points together and we have the LM curve.
The student should now answer the question:
(i) what happens if the Central Bank (the Bank of England)
increases the money supply?. Clue: The 45 degree line shifts outwards. Plot the
new LM curve and you should find it shifts outwards to the right but in
slightly an unusual manner.
The shape of the LM curve is a result of the
way we have drawn the speculative demand for money. It assumes that there is a
floor to the interest rate below which interest rates cannot fall and that
there is some minimum amount people will keep for speculative purposes (perhaps
zero).
What is the minimum you must know? The LM
curve is the combinations of R and Y which results in equality between planned
the demand for money and the supply. It slopes upwards. An
increase in the money supply shifts the LM curve outwards.
Analysing Government Economic Policy Using IS
and LM curves:
For
Figures referred to on this page click on:
http://staff.bath.ac.uk/hssjrh/MACFIG3.pdf
An increase in Government spending shifts the
IS curve outwards. There are several possibilities. In Figure 1a we get a
substantial increase in Y (GDP) and a small increase in R. What has happened?
The increase in G increases Y, this will cause
consumers to spend more money increasing C causing Y to increase still further
(the multiplier). This increase in Y results in an increase in the transactions
demand for money and with a fixed money supply this will leave less for
speculative purposes, driving up the rate of interest. (This views the rate of
interest as the ‘price of money’ if the demand increases with fixed supply, the
price (interest rates) rise. This increase in R will have a negative effect on
I, investment will actual fall thus reducing Y slightly from the boost the
increase in G and C had given it. We say that some I has
been ‘crowded out’ by the increase in G. In Figure 1b, the increase in Y
is much smaller and that in R much bigger than hitherto. What has happened? We
are further up the LM curve. To get money for transactions purposes with a
fixed money supply we need to shift money away from speculative purposes. Yet
on this almost vertical section of the LM curve speculative demand is close to
its minimum below which it will not fall. To get even a small amount switched
out of speculative demand results in a large increase in interest rates. But
this causes a reduction in I almost equal to the
increase in G. Crowding out is almost 100%.
In Figure 2 a reduction
in tax rates also ‘swivels’ the IS curve outwards. In Figure 2a there is a
substantial increase in Y and a small increase in R. What has happened? The
reduction in the direct tax rates increases disposable income,
this increases consumer expenditure increasing Y. Once more the multiplier
works to expand this initial increase in Y. This increase in Y results in an
increase in the transactions demand for money and with a fixed money supply
this will leave less for speculative purposes, driving up the rate of interest.
The analysis from here on in is the same as with the increase in G.
In Figure 3a, we have an increase in the
money supply shifting the LM curve outwards. There is once more a substantial
increase in Y and a reduction in R. What has happened? The increase in the
supply of money has reduced the rate of interest which has boosted investment.
This has increased Y and consumers have therefore increased their expenditure
(the multiplier) thus increasing Y still further. In Figure 3b, the increase in
the money supply has had no impact on either interest rates
nor Y. Why? Interest rates are already at their floor they cannot fall
any lower (the floor is probably about 1%-2% in practice). This is called the ‘liquidity
trap.’ It is possible that this is one of the problems facing
The 45 degree line diagram
For
Figures referred to on this page click on:
http://staff.bath.ac.uk/hssjrh/MACFIG45.pdf
The 2nd analytical format we will use to show
output determination is the 45 degree line diagram. This is simpler and less complete
than ISLM as it just looks at equilibrium in the goods market without focusing
on the money market. It does however have the virtue that it allows us to focus
on the workings of the multiplier. Figure 4 shows actual income on the
horizontal axis and planned expenditure on the vertical axis. Our consumption
function is the standard one we have used all along, b is the marginal
propensity to consume, t is the tax rate, hence (1-t)Y
is disposable income. Aggregate demand is found by adding to the consumption
line I+G. These are assumed to be fixed (exogenous)
That is AD = C+I+G. This tells us what for any level
of Y (actual income) planned expenditure (Yp=AD)
will be. Equilibrium is where planned expenditure = actual total income
(remember total income=total output = total expenditure, hence our definition
of equilibrium implies planned expenditure = actual expenditure. Right, now the
45 degree line in Figure 4 equates all points on both axes. Hence take an
income of Y1. this will result in planned
expenditure of Yp1 and because these two meet on the 45
degree line they are equal. Hence a level of I+G as shown will result in
equilibrium output of Y1. What happens then if I+G increases, say there is an increase in G to G’? Well if Y
stays at Y1 planned expenditure will be Yp3
where Yp3 ¹ Y1. Y1
is no longer an equilibrium level of income. What is? It will be again where
the new aggregate demand curve cuts the 45 degree line. This will be at Y2
where once again planned aggregate expenditure equals actual income.
Now notice that the increase in Y (Y2
- Y1 =D Y ) is greater
than the increase in G (G2 - G1 = D G). Why? Well as with ISLM, the increase in G increases Y, this will
cause consumers to spend more money increasing C causing Y to increase still
further (the multiplier). The ratio D Y/D G is called the multiplier. Let us briefly leave our diagramatical analysis and derive it algebraically:
Y=C+I+G (Closed economy (no imports or
exports), national income identity)
C = a + bYD
= a + b(1-t)Y
Insert the above into Y and we get:
Y = a + b(1-t)Y + I + G;
Y - b(1-t)Y = a + I + G ; Y(1-b(1-t)) = a + I + G
Y = a/[(1-b(1-t)] + I/[(1-b(1-t)] + G/[(1-b(1-t)]
Any increase in G (or for that matter I) will
be multiplied by 1/[(1-b(1-t)]. This is the
multiplier. If b equals 0.7 and t=0.4 then 1/[(1-b(1-t)]
= 1/[(1-0.7(1-0.4)] = 1/(1-0.42) = 1.72. Thus an increase in G of £1b will
increase Y by £1.72b. Note this keynes
argued, and all of this on IS-LM is pretty pure Keynesian policy analysis. That
not only would an increase in G increase Y by G it would increase it by more
than G. How much will this cost the Government? Well we have a tax rate of 0.4
and Y has increased by £1.72b, hence extra taxes raised equal £0.69b. The cost
to the Government will therefore have been just £1b - £0.69b = £0.31b.
[Although it may be even better than that, if the number of unemployed falls
and as a consequence, the amount paid in unemployment benefit were also to
fall]. Note too that for this we will have extra roads, school buildings, a
better health service, etc. It almost seems like having ones cake and eating
it.
See: http://staff.bath.ac.uk/hssjrh/multiplier[1].ppt
Figure 5. Also shows Keynesian economic policy at work. This
time a cut in direct tax rates swivels, the consumption function upwards,
increasing equilibrium output (Y) from Y1 to Y2. Note
again that although the Government has cut taxes, Y will increase and hence the
loss to the Treasury in terms of tax revenue will not be as great as might be
originally thought.
Keynes (on the right) with George Bernard
Shaw
We now turn to the third and final diagramatic apparatus we shall use to analyse output
determination, aggregate supply (AS) and aggregate demand (AD) curves. This is
supposed to be the newest of the three methods but the rationale dates back
almost to Keynes. The key difference between this and the other two methods is
that the price level is allowed to change. Let us first deal with aggregate
supply and repeat the rationale found in L+C (pp. 398-9). The Short run
aggregate supply (SRAS) curve shows the quantity of output that firms would
like to produce and to sell at each price level on the assumption that the
prices of all factors of production (inputs) remain constant.. The Long-run aggregate supply
curve (LRAS) does the same under the assumption that the prices of all
factors of production (inputs) have fully adjusted to any exogenous shifts
in the AD curve. We concentrate on the SRAS curve. The rationale I prefer is
that with fixed input prices (including wages) each firm will produce output at
the level at which marginal revenue equals marginal cost..
Now if the price of the firm’s output increases clearly the marginal revenue
for that product will increase. Hence the firm will increase output, declining
marginal returns means that as it does so marginal revenue will fall back. At a
higher level of output marginal revenue will again equate to marginal cost.. Hence the SRAS curve will be upward sloping,
higher prices will induce higher output.
For a graphical analysis of aggregate supply
and demand see;
http://www.whitenova.com/thinkEconomics/adas.html
For
Figures referred to on this page click on:
http://staff.bath.ac.uk/hssjrh/MACFIG67.pdf
The aggregate demand curve. This is equal to Aggregate expenditure: C+I+G+X-M. let us explore how an increase in the price level impacts
upon C, X and M. A change in the price level affects the wealth of assets
denominated in money terms in exactly the opposite way to how it affects the
wealth of those who issued it. We need to distinguish between ‘inside assets’
and ‘outside assets’. An inside asset is one that is issued by someone in the
private sector to someone else in the private sector (e.g. a bank loan). If it
loses value, one loses and one gains, but there should be no net change in
private sector wealth. Outside assets are issued by some agent outside the
private sector (typically government) to someone in the private sector. E.g. a pound coin. Now is the price level falls by 10% this
pound coin will buy 10% more than previously, the private sector holder of the
pound coin is richer and as a consequence consumers’ expenditure should
increase. This argument has been around for many years and originally put
forward by Pigou in arguing with keynes on the General Theory as a consequence it is
sometimes called the Pigou effect. How about the
effect on X and M? Simple really. If the price level
falls and prices in Germany (for example) do not fall or fall less than in the
UK and if the exchange rate stays the same, then British goods become more
competitive we export more and import less. Thus for all of these reasons
aggregate expenditure or demand increases as the price level falls (C
increases, X increases, M falls). Hence the AD curve slopes down as shown in
the diagram.
Where the curves cross we get the equilibrium
output and price level. Note the curves can shift for several reasons. Firstly,
as with the IS curve an increase in G will shift the AD curve outwards. Also a
change in the real exchange rate, say an increase, which makes domestic goods
less competitive, will shift the AD inwards. Similarly the AS curve can shift.
It tends to do so for two reasons. Firstly, technical
progress. This means that the same amount of labour working with the
same amount of capital (machines, etc) will produce more this year than last
simply because of technical progress, we become more efficient. This will shift
the AS curve to the right, as indeed will investment which increases the
capital stock. Thus a reduction in the real rate of interest which increases
investment may simultaneously shift the AD and the AS curves to the right.
However, there has been much made in the recent literature on ‘shocks’ which
can shift the curves. For example, in 1974 and 1979 we had the two oil price
shocks which raised input prices and thus raised marginal cost shifting the AS
curve inwards. Similarly there can be such shocks impacting on the AD curve and
as we shall see later a whole theory of ‘the business cycle may be built around
shocks and their after-effects.
Aggregate supply and demand curve analysis is
more recent than ISLM and the common view is that it is more relevant as it
relaxes the assumption that price are fixed. In fact little in it is new.
Keynes was aware shortly after he wrote the book of the Pigou
effect. The supposed impact on imports and exports of a change in the price level, depends upon the actual exchange rate not adjusting
to maintain the exchange rate constant which is problematical. Keynes too was
arguing for the use of expansionary fiscal (and also perhaps monetary) policy
at times when output was clearly stuck well below its equilibrium level, when
there was considerable excess capacity and when we are therefore .on the flat
part of the AS curve. There must be considerable doubt as to whether he ever
intended it to be used to ‘fine-tune’ the economy. Fine tuning is when the
Government changes the tax rate or Government spending or the rate of interest
with a view to increase growth by say 1% at a time when output is slightly
below its equilibrium level, or vice versa when it is slightly above this
level. And yet Keynesian policy has never been used since 1945 in such a way
(one can argue that
In addition there is a further problem in
that traditionally there have been three main ‘instruments’ of macroeconomic
policy. Government expenditure and taxation (which together form the tools of
fiscal policy) and the money supply or interest rates (the two are determined
simultaneously, set the money supply and you in effect set the interest rate
and vice versa). Now compare these three instruments to the accelerator, the
steering wheel and the brake when driving a car. It would be difficult indeed
for one person to have control of the steering wheel and another
the brake and the accelerator and still expect to get to your
destination safe and on time. But that in effect is what has happened in many
countries including the
http://www.bankofengland.co.uk/monetarypolicy/members.htm
for an informative article on fiscal policy:
http://oxrep.oxfordjournals.org/cgi/reprint/21/4/515 and test yourself: http://highered.mcgraw-hill.com/sites/0077099478/student_view0/chapter24/self-test_questions.html
Also see the following piece on the
Returning to the three sets of apparatus, my
own view is that none are entirely satisfactory. The 45 degree line illustrates
the multiplier at work. ISLM curves emphasise the interaction between the real
and the monetary side of the economy, but suffers from several disadvantages,
including its assumption that prices are fixed, whilst AD-AS curves have
several questionable assumptions (e.g. that relating to exports, imports and
the actual importance of the Pigou effect). I tend to
prefer ISLM, but it feels slightly dated. In truth none are satisfactory. The
economist works by simplifying a complex world into a model which he/she can
use to analyse that world. But restricting complex modern economies to just two
dimensions seems to me a simplification too far. Their role is at best a
teaching one, but one needs to be careful that in simplifying for teaching
purposes we do not mislead.
Exchange Rates
The exchange rate is the rate at which one
currency is exchanged for another, for example at the time of writing one can
exchange £1 for about US$1.69. This is just about one of the hardest of
economic variables to model amd to forecast. Most
economists would, reluctantly subscribe to the purchasing power parity theory
of exchange rates (PPP). Reluctantly because there is considerable doubt as to
how well it works. But without too much exaggeration it ‘is almost the only
game in tow’. Exchange rates are a crucial factor in impacting on imports,
exports and the balance of trade. It is therefore essential that the student
should know the PPP theory and also what the real interest rate is. However,
this will not be taught in the lectures. The lectures are a guide to the
course. They are not the course itself. This also includes the reading the
student is expected to do. To emphasise this point I do not lecture on it. You
should look up exchange rates, real exchange rates in the appendix in Lipsey and Chrystal and read and make your own notes.
However this is not enough. You should then go to and read Taylor and Taylor in
the Journal of Economic Perspectives. Fall 2004. Make
notes on it, in two pages you should have a succinct summary which serve the
place of lecture notes.
Inflation
Inflation is the percentage change in prices.
It is often misunderstood. It is not high prices, it is not a high cost of
living, it is the percentage change in prices. Now we
tend to measure prices by c consumer price index, in the
Over the years there have been two theories
of inflation which have dominated the literature. First there is the quantity
theory of money. This is based on the following IDENTITY (not an equation but
true by definition)
MVº PQ (1)
M=money supply (intuitively the number of
pound coins in the economy)
V=velocity of money (intuitively the number
of times an average pound gets spent)
MV= total amount spent in a given time period
P= average price of goods in transactions
T= the number of transaction, i.e. the number
of things bought and sold
PT= total amount spent in a given time period
Hence (1) is true by definition. Let us
rearrange
P º MV/Q (2)
Still an identity. But let us suppose T and V are constant such that
k=V/Q (3)
Hence
P=kM (4)
Now we have a theory, (3) is an assumption
not necessarily true. (4) tells us that if M increases
by 10% then P also increases by 10%. Hence the way to control inflation is to
control the money supply. The crucial issues are (i) are V and T really constant? and
(ii) can the Government, or the Bank of England, control the money supply. With
respect to (1) it does not matter too much if V and T do vary, provided (I)
they vary in a predictable fashion and (2) they do not vary in such a way that
P is insulated from an increased in M. For example, if an increase in M of 10%
simple reduced V by 10% then nothing would happen to prices. How might this
translate into an equation for a macoeconomic model?
Something like:
D Pt = b 1 + b 2D Mt + b 3D Pt-1 (5)
where D Pt-1 is as always
included to represent lags in economic behaviour. (Note D Pt is inflation in period t. D M is the
percentage change in the money supply. This is NON STANDARD notation,
generally D means the change in, not the percentage change.
The quantity theory of money is both an older
theory (Irving Fisher, Chicago, 1920s and others) and a newer theory (to a
considerable extent revived by Milton Friedman,
D Pt = f(Ut) (6)
That is it is shaped like Figure 8:
For
Figures referred to on this page click on:
http://staff.bath.ac.uk/hssjrh/MACFIG89.pdf
Problems emerged with the Philips curve in
the 1970s when we observed high inflation and high unemployment - something
which was not supposed to happen. Ed Phelps had put forward a theory in about
1970 (almost simultaneously with Friedman publishing a similar one) which
explained this. Basically it was argued that if workers expect prices to
increase by 10% then in order to maintain the attractiveness of the wage offer,we must add to (6) expected
inflation. Hence (6) becomes:
D Pt = f(Ut) + D Pte (7)
This was known as the expectations augmented
Philips curve. This is not the first time expectations have been mentioned in
this course, but in the 1960s and 1970s a lot of work was done surrounding how
people form their expectations. There are two main types of theories. Firstly,
that people form there expectations of what is happening in the future by what
has happened in the past (often termed a backward looking theory), "prices
rose by 20% last year, 15% the year above, they look like they are on a
downward trend I think they will rise by 12% this year). That is we extrapolate
from the past into the future. The most widely used form of extrapolative
expectations is adaptive expectations:
D Pet = a D Pet-1 + (1-a )D Pt (8)
That is current expectations of inflation are
a weighted average of expectations in the previous period and what actually
happened this period. Another way of writing (8) is that we adjust expectations
according to the previous error in expectations, hence
it is sometimes caused an error learning mechanism. In the mid late 60s early
70s Bob Lucas revived an idea in a paper by John Muth
that firms form their expectations according to the relevant economic theory
and applied it to expectations of inflation. Quite literally the rational
expectations theory implies that if the relevant economic is the quantity theory
of money, then workers, firms, housewives, everyone will base their
expectations of inflation according as to what is happening to the money
supply. This is often termed a forward looking theory.
Log on to Bank of England web site for video clip click on ‘what if’, choose windows media player broad band and you will get about a 20 minute video of Mervyn King, Governor of the bank, explaining monetary policy and indeed other things. http://www.bankofengland.co.uk/education/videos/index.htm
Also: http://research.stlouisfed.org/publications/review/07/01/Poole1.pdf and test yourself: http://highered.mcgraw-hill.com/sites/0077099478/student_view0/chapter26/self-test_questions.html
Unemployment and the NAIRU see:
http://wms.bath.ac.uk/live/Economics/unemployment.wmv
for a TV recording of this lecture which will not be
given ‘live’. As with all lectures it is examinable.
Many economic models have as a starting point
for unemployment that in order to produce Y(=C+I+G+X-M)
you will need so many workers (LD)
LD=f(Y)
subtract this number of workers from the Labour force (N) and
we have the amount of unemployment
U=N-LD
Thus one reason for unemployment is that
there may not be enough demand for workers. This is derived from output and
both may be expected to vary over the cycle which is why this form of
unemployment is called cyclical unemployment. Policy
Cure? Keynesian demand management using expansionary
fiscal or monetary policy? This is problematical, and currently out of
favour in many countries, partly because of the difficulties in such policy
which we have already analysed. But partly because of the
monetarist view of the world. ‘Monetarists’ is a relatively loose term
to describe a fairly wide range of views. Once more Milton Friedman is a key
player (Over the course of the twentieth century many of the leading
monetarists have come from the
OK, but if demand equals supply surely there
should be no unemployment and yet we no that in a modern economy there is never
any unemployment so how do the new classical school explain that one? In several ways. Firstly, they would argue that some people
will present themselves as being unemployed for the purpose of collecting
unemployment benefit. But in reality all of these are voluntarily
unemployed. Secondly there will always be some unemployment through people
leaving one job to go to another which suits them better, or leaving firms in
declining industries to move to jobs in expanding industries. The time in
between jobs is called ‘frictional unemployment’ of which more later. However, what if there are no jobs for the people in
declining industries to go to? This leads us to a concept known as structural
unemployment:
structural unemployment.
There are job vacancies and unemployed workers, but the vacancies are for
computer software engineers and the unemployed workers are ex-shipbuilders.
This is a skill mismatch. It may well be related to structural change in the
economy which sees some industries expand and others contract. In recent years
a separate term has been called technological unemployment, but in reality this
is no more than structural unemployment due to skill mismatches. Another
example would be where there are vacancies for office workers in
frictional unemployment
As we have seen even with no cyclical nor structural employment, there would
still be some unemployment as workers switch jobs and also workers enter and
leave the labour force. It takes time to obtain new workers. Exactly how much
time depends upon informational aspects of the job market and also how keen
workers are to accept unemployment. This in turn is related to the ‘replacement
ratio’ (RR, the ratio of benefits to pay when in work). An increase in the
replacement ratio will probably reduce the incentive for workers to accept any
job offer and to continue searching for a better job. It will therefore
increase unemployment.
Is this all there is to the new classical story?
Not exactly, even after allowing for people who are voluntarily unemployment,
for structural and frictional unemployment there still often appears to be some
unemployment which we cannot explain. This may be because wages are ‘sticky’,
e.g. in Figure 9 slow to move from W1 to W2 thus creating
equilibrium in the labour market. What can cause wages to be sticky in this
manner? There are a variety of possibilities: (i) overpowerfull trade unions who
prevent wages from falling for the benefit of their members in work (even
though it means that there are other workers who suffer from being without a
job because of this), (ii) implicit contract theory, briefly workers dislike
whilst firms are less risk averse. A very simple example: the risk averse
worker is willing to say to the firm, I know that in good times I could get
£300 a week, and in bad times only £250. I know too that on average there are
as many good times as bad times. Because I am risk averse I am willing to be
paid £270 in good times and bad. The firm bears the risk, but on average should
save on its wage bill. There is an ‘implicit’ contract between firm and worker
which prevents wages falling in a recession and clearing the market. (In L+C
this is dealt with on page 536 under the heading ‘long-term relationships’).
(iii) thirdly we look at the efficiency wage.and menu costs. Both of these are dealt with in L+C
(pp.536/7).
NAIRU - the Non Accelerating Inflation Rate
of Unemployment (a concept associated with the name of Steve Nickell. the Oxford University economist, but is very
similar to a much earlier concept known as the natural rate of unemployment
attributed to Friedman and Phelps). We made the point earlier that according to
the Philips curve inflation will depend upon the level of unemployment. Let us
go back to Figure 8 [the second of the two figures on the right hand side] in
the inflation lecture notes. We start with a situation where people have zero
expectations of inflation. A level of unemployment of U1 will then
lead to a rate of inflation of 2%. Regardless of the theory of expectation
formation we have if this level of unemployment is sustained people will come
to expect 2% inflation. In this case the level of inflation associated with U1
will now become 4% (2% because of the basic Philips curve + 2% inflationary
expectations). let is suppose unemployment continues
to remain at U1, people will then come to expect inflation to equal
4% and once they do inflation will increase to 6%. When will this stop? When
will inflation stop increasing? As long as unemployment remains at U1
inflation will increase without limit. The reverse argument holds for U2,
i.e. we will get steadily falling prices, deflation or negative inflation. Only
at U* will inflation be stable, neither increasing or decreasing.
This is the NAIRU.
The Costs of Unemployment and Inflation
The costs of unemployment are the costs to
the economy in foregone output, the personal costs (for example there is some
evidence that the health of the long-term unemployed worker and their family
suffer, as well as the impact on their income) and the impact on government
financing (generally overlooked, but it will push government finances into
deficit). The costs of inflation are more nebulous. There are
the "shoe leather" costs associated with people faced with rising
prices than expected searching for lower ones, the fact that with inflation the
value of pound coins will fall and people will cut back on their optimal
holding of cash. There will also be some redistribution of income from lenders
to borrowers and those on fixed incomes (e.g. pensions) will fall behind. But
we can overcome the latter, e.g. we can ‘index link/ pensions to the retail
price index (a measure of prices). More generally, however, inflation seems to
create a climate of uncertainty in which people (firms and consumers) are more
likely to take the wrong decisions. But before we get too carried away with
this, consider this in 1996 Korea had inflation of 5% in 1997 and growth in GDP
of 11% (in real terms). I suspect what many people are afraid of is
hyperinflation (e.g. in
The following links into a video – try real
player with Ed Phelps Nobel \prize winner.
http://nobelprize.org/nobel_prizes/economics/laureates/2006/phelps-interview.html
MACROECONOMIC MODELLING
http://wms.bath.ac.uk/live/Economics/Modelling.wmv
For a TV recording of this lecture which will
not be given ‘live’
also see;
The equations we have specified throughout
these lectures form the nucleus of a macroeconomic model such as that you have
been using on the IFS website. Let us specify a simple such model:
Ct = 100 + 0.6YDt +
0.3Ct-1 - 5D Pt
It = 50 + 4.0(Yt - Yt-1) - 10rt
Log(Xt) = 2.0 + 1.2log(WDt)
+ 0.8Log(PCt)
Log(Mt)
= 2.0 + 1.1log(Yt) - 0.8Log(PCt)
Yt = Ct + It + Gt
+ Xt - Mt
U = Log(Yt) - TREND
D Pt = 100 - 2Ut + 1D MSt
PCt = ERt*WPt/Pt
YDt = (1-t)Y
nt =
2.5 + 2.0Yt - 1.5MSt
rt = nt - D MSt
Let us briefly summarise these equations. Consumption
depends upon disposable income lagged consumption and declines with the
inflation rate. Investment depends upon changes in GDP (the accelerator) but
declines with the rate of interest (the opportunity cost of investment).
Exports increase with world demand (WD) and also with price competitiveness,
imports increase with GDP and decline with price competitiveness, We then have the national income identity. Unemployment
depends upon GDP and trend productivity. Inflation is a version of the expectations
augmented Phillips curve whereby expectations of inflation are equal to current
growth in the money supply (a form of rational expectations). Price
competitiveness is defined as equalling the exchange rate multiplied by the
price of world goods (WP) divided by the UK price of goods, Disposable income
is GDP multiplied by (1- the tax rate). The nominal interest rate (nt) depends upon the
depend for money (as proxied by Y, representing
transactions demand) and the supply of money. The real rate of interest equals
the nominal interest rate minus expected inflation (again equal to the current
growth in the money supply).
You will see we have 11 equations and
identities. The variables on the left hand side of these equations and
identities are endogenous to the model, they are
explained by the model. The remaining variables are exogenous. The model does
not explain these. There are two types of exogenous variables. WD, WP and TREND
are variables the government has no control over. However, G, MS, ER and t are
exogenous variables under the policy makers control.
We will call these the instruments of government policy. We will return to this
distinction later. For the moment we shall note that the first stage of using a
macroeconomic model is to specify the equations which enter the model and the
variables which enter those equations. These equations will be based on
economic theory, although clearly monetarists would come up with a different
model to Keynesians. This also involves deciding what to include within the
model (i.e. what is endogenous) and what to leave as exogenous.
The second stage is to estimate the model.
This will be done using econometric techniques and need not detain us here.
Just understand that we have ways of estimating the above coefficients (numbers
in the equations) based on, e.g., the way the consumption function has behaved
in the past. Having estimated a model it is then subjected to rigorous testing,
this is done on an equation by equation basis and again involves some
reasonably sophisticated econometrics, but it is also done on a complete model
basis. For example, if we estimate the model based on the period 1995 to 1994
and
Historical Simulation Ex post forecast Ex ante forecast
[____________________________________________________|_______________|_____________
1955 94 99 2004
use the model to simulate Y, C, I, U, etc over this same
period we are conducting a historical simulation. If the model fails to
simulate well over this period, then it is unlikely to forecast well and hence
would not be a good model. We can also evaluate how well the model forecasts
1995-99, when the outcomes are of course well known. This is a much more
difficult test than historical simulation and if it does well on this criterion
we may well have confidence in its ability to forecast the future. However, it
is often the case that although on strict statistical criteria each equation is
excellent, when combined together into a model they do not perform well.
Having constructed a model we might use it
for a number of purposes. Firstly, to forecast the future, i.e. engage in
ex-ante forecasting. However, before we can forecast 2000-2004, we will need
values for the exogenous variables. This means we will have to make some
assumptions about world demand, world prices and the trend level of output. We
also need to make assumptions about the values for the Government’s policy
instruments, i.e. values for G, t, MS and ER. Of course if it is the Government
or the Bank of England making the forecast these assumptions should be pretty
accurate, but even we can make pretty good guesses on the basis of the
Government’s stated intentions. Is this ‘all’ there is to forecasts? Not quite,
when you look at the forecast value for GDP or inflation you may well feel they
are not quite right. You may have a hunch as to what will happen, you may be aware that other factors are saying different
things, or the Chancellor of the Exchequer may simply tell you that you
forecast for GDP is too low and inflation too high. What can you do about it?
You can engage in ‘con adjustment’. For example, if you change the constant
term in the consumption function from 100 to 105 this will not only increase C,
but feed through into Y, I, U, etc. Similarly you can reduce the constant in
the inflation equation from 100 to 98 and inflation will fall. The name ‘con
adjustment’ is unfortunate in its connotations, but it stems from the fact that
typically we would adjust the constant terms in these equations. You then take
the new forecasts back to the Chancellor who will be very pleased with you.
Every single forecast you every see from a sophisticated macro model has been
con adjusted. Every one, we all do it. The defence is that by and large the
process does improve model forecasts. But does not invalidate the whole process
of macro-modelling? Perhaps, my own view is that the model provides information
about what is likely to happen which may well be at variance with your
expectations as a results you forecast somewhere in between. If this proves
better than your original expectations, the model has done its job. Remember
too the Treasury model may have some 600 odd equations, but these have to model
the complex interactions of 55 million people plus thousands of firms.
But models can also be used in policy
analysis. You for example have experimented with changing some of the
‘instruments of government policy’ with a view to examining their impact on
some of the ‘targets’ (growth, unemployment, inflation, balance of trade). The
Treasury do this also with their budget and expenditure plans. The Bank of
England too have a model of the economy, they will be putting the budget
changes through this model with a view to what will be its effect on inflation.
If it raises it above their 2.5% target then the Monetary Policy Committee will
raise interest rates accordingly. The techniques also exist for models to give
answers to the question what should interest rates, taxation and G be to
maximise welfare. You will have to provide information in the form of what, for
example, you wish unemployment and inflation to be. One can also use models to
forecast for example the impact of us joining the Euro, or alternatively of us
leaving the EU. But remember the more unusual the question you ask of a model,
the more removed from past experience, the more uncertain will be the answers
the model gives. It is best used for looking at the impact of small changes in
policy instruments over the medium term (say five years). Note too that the
techniques we have illustrated here could be easily transferred to building a
model to forecast sales of a particular product for a firm, or the impact of a
price change or an advertising campaign on profitability.
See: http://news.bbc.co.uk/2/shared/bsp/hi/pdfs/09_04_08_imf_chapter_2.pdf
SUPPLY SIDE ECONOMICS AND BANKRUPTCIES
You will note that the model we specified is
Keynesian in the sense that it is centred around the
GDP identity. GDP equals the sum of Consumers expenditure, investment
expenditure and so on. This then gives us total output,
this determines unemployment which feeds through into inflation and so on. Most
models are of this type. Monetarist models tend to differ from Keynesian ones
not in the structure of the model, but in its properties. For example, in A
Keynesian model the multiplier will be much greater than the monetarist model
where its long-run value may well be zero. We do not have to model things this
way. We could take the alternative approach - which is probably the way people
used to think of things prior to Keynes - of what is the economy capable of
producing? This then is divided up between consumption, investment etc. A few
models have done this, they have had some impact on
our theoretical understanding of the economy but little real impact on
macroeconomic modelling. There has been even less research done on what we
might call disequilibrium models. Let me illustrate this. The national income
identity now determined the total level of demand (YD) in the
economy:
YDt = Ct + It + Gt
+ Xt - Mt
Potential output (YSt)
is given by a production function linking output to the labour force and the
capital stock:
YSt = f(K,L)
Actual output is equal to the minimum of the
two:
Y = min(YS,
YD)
If YD < YS then we
are in the Keynesian world. The possibility that YD > YS
is rarely analysed. At best the statement would be that we are in the
vertical section of the aggregate supply curve, with inflation and FULL
EMPLOYMENT. The possibility that YD > YS with
significant levels of unemployment is just not considered. But let us examine
what happens in a recession. Output falls, unemployment rises, profits fall and
some firms go bankrupt and other firms cut back on there operations. Either way
capital is likely to be scrapped. We have seen this many times with the decline
of the coal industry, the collapse of shipbuilding, the loss of the motor bike
industry.
A simple model illustrates this:
YS =
10K
L = 0.5Y
YD
=2000
Y = Min(YD,
YS)
L = 0.5Y
With K (capital stock) = 200; YD
=2000, the economy is in equilibrium with aggregate supply equal to aggregate
demand and L (the number of workers in employment = 1000. In this particular
economy that is all the workers there are. Now let us suppose a ‘shock’ hits
the system and reduces aggregate demand to 1800. If wages are fixed (i.e.
‘sticky’ to use the prevailing jargon) then L will fall to 900 and 100 workers
will be unemployed. The Keynesian remedy is to increase Government spending or
cut taxes in a bid to boost YD back up to full employment level at
2000. (The monetarist remedy is to try and change the labour market so as to
make wages less ‘sticky’). But there is a potential problem with the Keynesian
solution what if the recession reduced profits and this led to an increase in
bankruptcies and the loss of capital stock? Let us suppose this
results in capital stock falling from 200 to 190. YS
now equals 1900. Keynesian policies will work as long as they seek to
boost output between the range 1800 to 1900, beyond
this range, they will be pretty ineffective. Within this range YD
> YS, several things are likely to happen (i)
inflation will rise as the demand is greater than what the economy can produce,
(ii) the balance of trade will worsen as unsatisfied demand for domestic goods
is made good by imports and (iii) profits will increase and there will be a
boost to investment (both by existing firms and new company starts). Over time
this increased investment will boost YS back up to 2000 where full
employment will be restored, but in the meantime we will have inflation and a
balance of payments deficit. This perhaps is the story of Keynesian policy and
the economy in
But how can it explain why the
But the legislation is controversial.
Bankruptcies perform several conflicting roles within the economic system.
Firstly, they are the engine of Schumpeter’s process of creative destruction by
which the old (and inefficient?) give way to the new (and efficient?), thus
paving the way for economic process. Resources are redistributed within the
economy within an optimal manner towards industries where there marginal productivity
will be greater. In this way we make progress. Bankruptcy procedures also
perform the role of allowing the failing firm’s creditors to recover their
money. This is an essential part of the incentive structure in a capitalist
system, firms and individuals are encouraged to ‘lend’ money (often implicitly
by, e.g. paying for goods in advance of their delivery or alternatively
supplying products in advance of payment) in the knowledge that if things go
wrong there is a process open to then to retrieve their money. An ‘optimal ‘ bankruptcy system must seek to balance these differing
perspectives. It must provide for the efficient ‘disposal’ of firms in
declining industries or the simply inefficient entrepreneur, whilst
simultaneously minimising the loss of potentially viable firms. In addition it
must provide for the retrieval of as much of creditors money as quickly as
possible. Clearly many of these are conflicting objectives and the optimal
bankruptcy system must seek a compromise between these different objectives.
One final point to note. The term ‘supply side economics’ is a term associated
with right wing economists - frequently monetarists, who argue that low taxes
are distortionary and, e.g. a cut in income tax would
stimulate labour supply and effort - in effect shifting the aggregate supply
curve to the right. The same is true of, e.g. reducing taxes on firms, which
would stimulate investment, increase capital stock and again shift the
aggregate supply curve to the right. The term is also associated with those who
would, for example, cut unemployment benefits in an effort to increase the
incentive to work and increase the flexibility of the labour market. Clearly,
even ignoring the arguments I have put above this is a relatively limited view
of what impacts on the supply side of the economy. For example, education which
increases labour productivity will also shift the aggregate supply curve to the
right.
See: Does
Chapter 11 Reorganisation Remain a Viable Option for Distressed Businesses for
the Twenty-First Century? By: Miller, Harvey R.; Waisman,
Shai Y.. American Bankruptcy Law
Journal, Spring2004, Vol. 78 Issue 2, p153, 48p; (AN 14743936)
Also: In The
http://en.wikipedia.org/wiki/Chapter_11%2C_Title_11%2C_United_States_Code
In The
http://en.wikipedia.org/wiki/Insolvency_practitioner
Also:
http://en.wikipedia.org/wiki/Receivership
Also look at the following paper, the
introduction, the section entitled “American Bankruptcy and Chapter 11’ and the
conclusion.
http://www.springerlink.com/content/w117285377164j86/fulltext.pdf
ECONOMIC CYCLES.
Terminology is inexact. When people talk of
cycles they often mean the economy booms and then the economy goes into recession
rather than follows an even path across time. Thus the causes of ‘the cycle’
are the causes of the ‘boom and bust economy’ (which you will note that every
Chancellor of the Exchequer since Selywn Lloyd in the
early 1960s has ‘successfully’ eradicated). Strictly speaking however, a cycle
is more than that. It is the regular nature of occurrence of booms and
recessions that make a cycle. Thus a four or five year cycle which constitutes
what is commonly thought of as the business cycle means that EVERY five years
growth will peak. Now clearly this is very much more than just boom following
bust in a random manner. And leads to consider issues such as what could cause
such a regular occurrence. What are the possible causes.
1. Political: At the moment probably the
dominant cause of the cycle is political linked to elections What
determines the outcomes of elections? "the
economy, stupid". With this dictum in mind Governments try and steer the
economy so that at election time, as much as possible of the following is true,
unemployment low and/or falling, inflation low, growth high, incomes rising,
taxes falling and public services improving. The exact mixture between falling
taxes and rising expenditure on public sector spending varies between left and
right wing Governments. But typically about two years before an election, the
economy will receive a fiscal boost from the Government in the form of
increased G or falling taxes. This is frequently augmented a year later. Then
once the election is safely out of the way, Governments will reverse this
policy, partly out of economic necessity as the economy overheats,
inflation and the balance of payments begin to worsen. In general governments
do the nasty, unpopular things early in their term of office. But they will
also make savings so as to create a ‘war chest’ which can be given away in time
for the next election. Now being as elections are about every four years in
many countries, but in particular in the
2. Economic: Systematic Shocks. All the
components of GDP have a stochastic (random) component. One theory of the cycle
is that a negative shock hits for example consumers
expenditure, there is some sort of adjustment process which generates a cycle.
For example, a negative shock in 2000 to consumers
expenditure, firms will respond with a lag by reducing output and laying off
workers, investment will decline which calls forth a further reduction in
output. Thus even if the shock is a one off with no further impact on consumers
expenditure 2001 may still see a reduction in GDP below its equilibrium level,
only over time and with no further shocks will this equilibrium be restored.
However in 2002 there may well be another shock which once again knocks the
economy off course. There is probably something to this theory, but I remain
sceptical about how RANDOM shocks can produce REGULAR movements in GDP. Sure it
can explain booms and recessions but NOT regular booms and recessions every
four to five years.
2. Economic: Accelerator Multiplier
Interaction. There is a permanent shift in GDP say from 1000 to 1100 which
means that capital stock should increase from 500 to 550. However as investment
is in itself part of GDP this increases GDP too an increases exacerbated by the
multiplier so that GDP rises to 1200 which calls forth another increase in
investment which is also exacerbated by the multiplier so that GDP now stands
at 1220 and capital stock has increased to 610. At this point output has hit a
ceiling, no further output is possible. Hence investment because of the
accelerator falls to zero and output falls back towards its equilibrium level
of 1100. But to produce this we simple need a capital stock of 550, whereas in
reality capital stock has increased to 610. Faced with excess capacity firms
will cut back on replacement investment (investment which replaces worn out
machines etc, but does not add to capital stock). This too is part of GDP which
hence will remain below its new equilibrium level of 1100 until capital stock
has once again fallen to 550. This is illustrated in Figure 27.2 of L+C but the
diagram is wrongly labelled it should say equilibrium real GDP not just real
GDP.
See (particularly Hicks’ trade cycle after
Figure 1): http://homepage.newschool.edu/het//essays/multacc/multacc.htm
The business cycle is not the only cycle
economists have studied. Early research focused on an eleven year cycle which
was linked to ‘sun spots’. These are spots on the sun which coincide with extra
large periods of emissions of solar flares and x and other rays. They have a
regular 11 year cycle and were thought to influence agricultural conditions at
a time when agriculture was much more of a driving force in the economy than it
is today. We shall be concentrating on a possible 50 year cycle linked with the
Russian economist ‘Kondratief’. Working in the 1920s
he suggested the existence of such a cycle (or long wave) by looking at data
for
Schumpeter’s explanation of the business
cycle is relatively straightforward. It centres on the concept of innovation, a
change of first magnitude (electricity, the railways, the
microchip). Prior to this the economy should be in a state of near-equilibrium,
where the economy has seen few recent innovations and entrepreneurs are
beginning to be put under pressure as the expansion of old markets has dried
up. Hence they will be receptive to new ideas (note the innovation may have
been invented years even decades before, but it needs entrepreneurs to take it
up and this they will do when the profits are under pressure). This primary
wave of innovation is succeeded by a secondary one brought about by induced
purchases of inputs and consumer goods. The upswing ends as the potential for
exploiting the new innovations dries up and also because many enterprises begun
in the secondary wave may not be viable in the long-run. The downswing may
simply take the economy back to a new equilibrium in which old patterns of
production have been replaced by new ones. An inevitable part of this is the
disappearance of old firms and the transfer of their resources to new ones. A
process he called ‘creative destruction’ and which he thought essential to the
workings of the capitalist system. He was therefore against government action
to stop the contraction of old industries such as tariffs, subsidies or cheap
loans, even though there is the possibility that in the downturn the economy
might overshoot equilibrium and go into recession (too many unviable firms get
started in the secondary wave and there demise pushes the economy into recession).
Thus the first part of the cycle is a wave of entrepreneurial activity built on
a cluster of new innovations. New firms being attracted by the high profits
being earned by the innovators are attracted into the industry. But there is
soon overproduction and capacity. Many firms will have borrowed heavily and
excess capacity drives prices and profits down, they will experience
difficulties, perhaps even becoming bankrupt.
Schumpeter had little to say about
unemployment and wages as such. But his policy recommendations to deal with
business cycles can be found in The Theory of Economic Development. The
most important remedy in the long-run was the greater familiarity of
businessmen with the working of the cycle. Foreseeing what is likely to happen
in the future helps them to take steps to neutralise cycles. In the short-run
he also favoured Keynesian-type government expenditure policies whereby new
construction by government enterprises is postponed to periods of depression. (note he was writing in 1911, pre-Keynes). With respect to
monetary policy he was against an indiscriminate and general increase in credit
facilities which he linked with inflation, but also hindered the closure of
firms in declining industries. However, he did favour a selective monetary
policy aimed at differentiating between those firms made obsolete by the
innovations from other firms, in trouble for a variety of temporary reasons,
but not fundamentally unviable.
SEE:
http://innovationzen.com/blog/2006/07/29/innovation-management-theory-part-1/
http://en.wikipedia.org/wiki/Kondratieff_Cycle
ECONOMIC GROWTH
http://staff.bath.ac.uk/hssjrh/growth.ppt
Macroeconomics is not just about one country at one point in time. If it is valid, its is applicable to all countries. We now look at developing countries.
THE IMPACT OF AID ON GROWTH AND POVERTY
Aid in its modern form has been with us since 1950. Its purpose has been to
lift countries out of poverty and propel them on the path to a stage when they
do not need aid. That is they are n a path of self-sustaining growth and out of
poverty.
For video on aid See: http://www.dfid.gov.uk/Media-Room/Sights-and-sounds-gallery/education-tanzania/
also see:
and look on ‘the world bank video clip and also the
Reference: Symposium on Aid in June 2004 edition of the Economic Journal.
MHV - Mosley Hudson and Verschoor paper; CD - Collier Dollar paper and DHT Dalgaard, Hansen and Tarp
How does aid work?
Take the production function: Y =f(K,L)
Theoretically perhaps the main impact of aid, at least in the early years, is to increase capital stock K and from this output increases and hence so do living standards. Aid could also impact on the quality of L (the labour force) via funding education. But aid is intended to be more effective than this. When a country is very poor – with large numbers of people earning less than a dollar a day – the resources are not really available to finance significant amounts of investment from savings nor for governments to finance public goods and infrastructure from taxation. By providing aid, increasing K and GDP it is hoped that this will change that countries will be able to fund public and private investment from their own savings and taxation.- self sustaining growth in the words of Rostow back in the 1950s.
Food aid, health aid, emergency aid Impacts on price mechanism Corruption fungibility
Empirical work
The debate on the impact of aid on the economies of developing countries is a rapidly evolving one. Early studies by, e.g., Papanek (1973) reported a positive impact of aid on growth within a multiple regression context. This conclusion allowed policy makers to entertain the possibility that poverty across the world could be largely eradicated. Unfortunately after some forty years the poor are still with us, poverty shows relatively little signs of disappearing and until recently enthusiasm for aid amongst donors had declined.
A number of economic studies help explain and clarify this apparent failure. Mosley, Hudson and Horrell (1987) found it impossible to establish any significant relationship between aid and the growth rate of developing countries. They suggested that this might be because of the possibility of leakages into non-productive expenditure in the public sector and the transmission of negative price effects to the private sector. The former is the problem of fungibility. Aid intended for some capital infrastructure project gets spent on the military, or cars for government ministers or even tax cuts. Perhaps even more worrying is high level corruption. To put it at its simplest aid goes into the pockets of the ruling elite. More on this is given later
The price effects? A very simple example. A large construction project funded by aid pays workers high salaries. Farmers leave their farms attracted by the high wages. The project lasts four years. Four years of good wages, but at the end of this the farm has deteriorated and is difficult to grow crops on. Another example, if aid is used to subsidise bread and other food products, this keeps their price ‘low’ and reduces the incentives for home producers.
More recently, however, work by
Burnside and Dollar (2000) (who were then at the World Bank) was to put forward
a more optimistic view of aid effectiveness. They concluded that aid had a
positive impact on growth for those developing countries with good fiscal,
monetary and trade policies in place, but had little impact for those countries
who were following poor policies. This therefore,
partially at least, provided an explanation of why aid had been found to have
little positive impact on growth in previous empirical work. It also provided
specific criteria for targeting aid. This was then built upon by Collier and
Dollar (2001 and 2002) (Paul Collier was then at the Bank, and later moved to
This work has had an extraordinary impact upon policy and Easterly (2003) as well as Dalgaard, Hansen and Tarp document how it has influenced both individual governments and international organisations – including the American government. Once more the eradication of poverty is a goal to which donors can and do enthusiastically commit and for this reason alone their work is of great significance.
However, it has not been without its critics. For example, Hansen and Harp (2001), have argued that the inclusion of an aid squared term, implying diminishing returns to aid, neutralises the significance of the aid-policy interaction term (i.e. aid does work and it works whether or not combination with good policy, but its effect is nonlinear). Dalgaard and Hansen (2001) also conclude aid’s positive impact on growth is characterised by diminishing returns (i.e. nonlinear again, a moderate amount of aid goes a long way, double it and its not so effective). Similarly Easterly argues that the Burnside-Dollar conclusions are not robust with respect to alternative, "equally plausible" definitions of aid, policies and growth. This is important as if the criticisms are valid, aid allocation rules based on the importance of good policies will unfairly penalise some countries.
As an aside this characterises applied economics. We have limited amounts of data to work with and are trying to explain a complex world with just a few variables. Slightly change an equations specification and results can change.
Good policies
The first issue to consider in this context is what constitutes good policies. Burnside and Dollar began with a vector of variables, inflation, unemployment and trade openness which are basically macro focussed and probably endogenous with respect to growth at the very least. CD partially recognising the limitations of restricting good policies so narrowly make use of the CPIA, as in their earlier papers (Collier and Dollar, 2001 and 2002). This is a World Bank measure which assesses a country’s policy and institutional framework in twenty dimensions, including e.g. economic management and policies for social inclusion. Hence compared to the original Burnside Dollar this is wide ranging indeed and undoubtedly a significant step forward.
A key element of the debate is whether either conditionality or selectivity is the appropriate form of linkage of aid to policy. Conditionality is the giving of aid to developing countries on the basis of promises to follow good policies in the future. CD argue that such promises are seldom kept and that, e.g. work by Alesina and Dollar (2000) finds no relationship between official finance and policy reform. They argue that aid donors should base their allocations upon good policies being in place now, rather than promises that they will be in the future.
I have always believed – unlike others in the debate - that the differences between selectivity and conditionality are superficial. The one says you get aid if you promise to be a good boy or girl in the future pursuing proper policies. The other says you get aid if you have pursued good policies in the past. But both are supposed to reward governments who follow good policies and thus provide an incentive for them to do so. Conditionality offers the advantage that it can by negotiation tailor appropriate policies to a countries needs. But it needs to be properly policed with punishment for countries who renege on promises. Selectivity is strong on enforcement – you only get aid if you already have good policies in place, but is a little less flexible in terms of tailoring policies to meet specific country needs. In addition it ignores the possibility that good policies may be facilitated by aid, i.e. aid may be a precondition for the implementation of good policies.
The debate has not been without a certain amount of heat and passion – rightly as this is a critically important issue. That should not, however, disguise the considerable amount of consensus which now exists on the central question of whether aid impacts positively in reducing poverty. The answer is in some circumstances, YES. Different economists simply disagree on the circumstances.
The Economist’s Dilemma
Aid comes in many forms including orientated towards health, food and emergency aid (e.g. Tsunami). Not all of these will impact upon GDP and poverty as indicated above. They are not meant too. But paradoxically there may be adverse affects. Take expenditure on health which reduces infant mortality. More children live to adulthood, but without a reduction in the number of children per family one is simply expanding the population beyond what the economy can support and quite possibly condemning the country to permanent poverty. Economics is a hard science. The figures below illustrates this.
Figure 6 shows that in low income countries as a whole the age
dependency ratio [the ratio of those not of working age to those who are] rose
in the period 1960-1971 although it has since fallen back to below the level
in1960. But for (sub-Saharan
A More in Depth Analysis of Why Aid Does not Work as well as it should.
Fungibility
Fungibility is the switching of aid money into non-productive uses, typically the provision of rents to politically powerful people (Feyzioglu et al. 1995). The good effect of aid-financed projects (the main micro result) was counterbalanced, in a number of ‘less open’ economies, by switching of aid into wasteful consumption expenditures to gratify the preferences of rent-holders.
Fiscal and other management
There is a risk that aid may erode
the tax base, It is a form of revenue, hence why try
to raise money from other sources? indeed Moore (1998)
and Brautigam (2002) demonstrate a negative
correlation between aid inflows into a country and the size of its tax base. Of
course, the pattern of causation cannot be inferred from a mere correlation,
and it is clear that there are some African countries which have escape from
the vicious circle, notably
Corruption, governance and state fragility
During the 1990s, it became clear that the failure of poor, ‘non-convergent’ developing countries to grow or reduce poverty was often associated with the failure of governments and associated institutions to function in a just and transparent way. The role of aid broadened from simply the improvement of efficiency and equity to the improvement of governance, and aid effectiveness must now be judged not only in terms of its ability, in the short term, to improve growth performance and reduce poverty, but also in terms of its ability to improve the performance of governmental and non-governmental institutions over the long term. Once more this is linked to the concept of conditionality and opens up the way for not only aid effectiveness to be improved as it is used in the way donors intend, but from a further knock-on effects from the improvement in governance.
In an important paper focused on Africa Brautigam and Knack (2004) find that in high aid-dependent countries aid provides more opportunities for, and thereby increases the level of, [the Transparency International index of] corruption and also reduces tax effort. It thus has a long-term depressive effect on institutional quality. If these findings are correct, they imply that what appears now to be a positive short-term effect of aid on growth and even on poverty is being undermined by long-term forces of institutional erosion which could cause the short-term effect to be unsustainable and once more brings into question the whole rationale for aid.
Absorptive Capacity Problems
Absorptive capacity potentially affects all developing countries. De Renzio classifies the constraints under several headings: (i) Macroeconomic Constraints. Sudden positive increases in aid inflows can cause an appreciation of the currency, damaging exports, i.e. Dutch disease, (ii) Institutional and Policy constraints. Aid requires expertise and organisational infrastructure on the part of the recipient to be able to administer it effectively. These are often in short supply in many recipient countries and a sudden increase in aid can stretch them beyond the limit and (iii)Technical and managerial Constraints. Other forms of human capital are also frequently in short supply. This is evident in for example, the number of doctors, teachers, nurses, administrators and managers.
Donor behaviour too must share part of the responsibility for aid’s relative ineffectiveness. In this section we shall examine this.
Donor Coordination
As noted by Brautigam and Knack(2004, p. 262) ‘numerous agendas
and projects, with numerous donors, can render ineffective any government’s
efforts to manage its aid resources’. Roodman (2006b)
focuses on problems arising from the recipient’s administrative burden. He
concludes that when projects proliferate beyond a certain point, the effective
marginal utility of aid declines sharply, and can even become negative. There
are also the simple transaction costs of dealing with a large number of donors.
In 2002 there were 25 official bilateral donors, 19 official multilateral
donors and about 350 international NGOs operating in
Donor Consistency
It is fairly clear from the micro-literature that, other things being equal, instability in aid flows tends to depreciate its quality. This confronts aid donors with a fundamental dilemma, since the ‘reactive’ forms of aid have been introduced, precisely in order to improve the quality of aid by making it more performance – and need-responsive. In other words there is a trade-off: if the donor seeks to improve aid quality by reducing volatility, most ways of doing this – e.g. reducing the penalty for poor performance on budget support aid – will reduce the donor’s flexibility. An attempt has been made to finesse this dilemma by Gelb and Eifert (2006), who propose that, in the event of a negative shock to aid disbursements caused for example by a fall in the recipient’s CPIA rating, this should be automatically counteracted by drawings on a buffer fund, which would stabilise aid disbursements.
Aid Volatility
There has been some recent work looking at aid volatility and its potential impact on aid effectiveness. The idea is that a steady stream of aid of, e.g. 2007: 5%; 2008: 5%; 2009:5% is more effective than more volatile aid flows: e.g. 2%; 12%; 3%. In the following paper we establish that this is indeed the case and in addition downside (negative) volatile (aid shortfalls from the average) are more destructive than upside (positive) volatility. The former can cause the cancellation of projects. The latter can cause absorption problems, i.e. they have difficulty in using the aid in a constructive manner. There are also problems for aid recipient countries in the number of donors it may have to deal with in a country, aid donors such as DFID (UK) and also the World Bank, but also NGOs. This takes time and effort and uses up scarce bureaucratic resources.
The following regression table illustrates this (taken from the paper referenced at the end of this section.
The debate on the impact of aid on the economies of developing countries is a rapidly evolving one. Early studies by, e.g., Papanek (1973) reported a positive impact of aid on growth within a multiple regression context. This conclusion allowed policy makers to entertain the possibility that poverty across the world could be largely eradicated. Unfortunately after some forty years the poor are still with us, poverty shows relatively little signs of disappearing and until recently enthusiasm for aid amongst donors had declined.
A number of economic studies help explain and clarify this apparent failure. Mosley, Hudson and Horrell (1987) found it impossible to establish any significant relationship between aid and the growth rate of developing countries. They suggested that this might be because of the possibility of leakages into non-productive expenditure in the public sector and the transmission of negative price effects to the private sector. The former is the problem of fungibility. Aid intended for some capital infrastructure project gets spent on the military, or cars for government ministers or even tax cuts. Perhaps even more worrying is high level corruption. To put it at its simplest aid goes into the pockets of the ruling elite. More on this is given later
The price effects? A very simply example. A large construction project funded by aid pays workers high salaries. Farmers leave their farms attracted by the high wages. The project lasts four years. Four years of good wages, but at the end of this the farm has deteriorated and is difficult to grow crops on. Another example, if aid is used to subsidise bread and other food products, this keeps their price ‘low’ and reduces the incentives for home producers.
Boone (1996) also concluded that aid does not significantly increase growth nor benefit the poor. It did however increase the size of the government.
More recently, however, work by
Burnside and Dollar (2000) (who were then at the World Bank) was to put forward
a more optimistic view of aid effectiveness. They concluded that aid had a
positive impact on growth for those developing countries with good fiscal,
monetary and trade policies in place, but had little impact for those countries
who were following poor policies. This therefore,
partially at least, provided an explanation of why aid had been found to have
little positive impact on growth in previous empirical work. It also provided
specific criteria for targeting aid. This was then built upon by Collier and
Dollar (2001 and 2002) (Paul Collier was then at the Bank, and later moved to
This work has had an extraordinary impact upon policy and Easterly (2003) as well as Dalgaard, Hansen and Tarp document how it has influenced both individual governments and international organisations – including the American government. Once more the eradication of poverty is a goal to which donors can and do enthusiastically commit and for this reason alone their work is of great significance.
However, it has not been without its critics. For example, Hansen and Harp (2001), have argued that the inclusion of an aid squared term, implying diminishing returns to aid, neutralises the significance of the aid-policy interaction term (i.e. aid does work and it works whether or not combination with good policy, but its effect is nonlinear). Dalgaard and Hansen (2001) also conclude aid’s positive impact on growth is characterised by diminishing returns (i.e. nonlinear again, a moderate amount of aid goes a long way, double it and its not so effective). Similarly Easterly argues that the Burnside-Dollar conclusions are not robust with respect to alternative, "equally plausible" definitions of aid, policies and growth. This is important as if the criticisms are valid, aid allocation rules based on the importance of good policies will unfairly penalise some countries.
As an aside this characterises applied economics. We have limited amounts of data to work with and are trying to explain a complex world with just a few variables. Slightly change an equations specification and results can change.
The first issue to consider in this context is what constitutes good policies. Burnside and Dollar began with a vector of variables, inflation, unemployment and trade openness which are basically macro focussed and probably endogenous with respect to growth at the very least. CD partially recognising the limitations of restricting good policies so narrowly make use of the CPIA, as in their earlier papers (Collier and Dollar, 2001 and 2002). This is a World Bank measure which assesses a country’s policy and institutional framework in twenty dimensions, including e.g. economic management and policies for social inclusion. Hence compared to the original Burnside Dollar this is wide ranging indeed and undoubtedly a significant step forward.
MHV are more concerned with the issue of whether growth automatically benefits the poor or whether aid could be targeted at the poor in what is termed as ‘pro-poor expenditure’.
A key element of the debate is whether either conditionality or selectivity is the appropriate form of linkage of aid to policy. Conditionality is the giving of aid to developing countries on the basis of promises to follow good policies in the future. CD argue that such promises are seldom kept and that, e.g. work by Alesina and Dollar (2000) finds no relationship between official finance and policy reform. They argue that aid donors should base their allocations upon good policies being in place now, rather than promises that they will be in the future.
I have always believed – unlike others in the debate - that the differences between selectivity and conditionality are superficial. The one says you get aid if you promise to be a good boy or girl in the future pursuing proper policies. The other says you get aid if you have pursued good policies in the past. But both are supposed to reward governments who follow good policies and thus provide an incentive for them to do so. Conditionality offers the advantage that it can by negotiation tailor appropriate policies to a countries needs. But it needs to be properly policed with punishment for countries who renege on promises. Selectivity is strong on enforcement – you only get aid if you already have good policies in place, but is a little less flexible in terms of tailoring policies to meet specific country needs. In addition it ignores the possibility that good policies may be facilitated by aid, i.e. aid may be a precondition for the implementation of good policies.
The debate has not been without a certain amount of heat and passion – rightly as this is a critically important issue. That should not, however, disguise the considerable amount of consensus which now exists on the central question of whether aid impacts positively in reducing poverty. The answer is in some circumstances, YES. Different economists simply disagree on the circumstances.
The Economist’s Dilemma
Aid comes in many forms including orientated towards health, food and emergency aid (e.g. Tsunami). Not all of these will impact upon GDP and poverty as indicated above. They are not meant too. But paradoxically there may be adverse affects. Take expenditure on health which reduces infant mortality. More children live to adulthood, but without a reduction in the number of children per family one is simply expanding the population beyond what the economy can support and quite possibly condemning the country to permanent poverty. Economics is a hard science. The figures below illustrates this.
Figure 6 shows that in low income countries as a whole the age dependency
ratio [the ratio of those not of working age to those who are] rose in the
period 1960-1971 although it has since fallen back to below the level in1960.
But for (sub-Saharan
A More in Depth Analysis of Why Aid Does not
Work as well as it should.
Fungibility
Fungibility
is the switching of aid money into non-productive uses, typically the provision
of rents to politically powerful people (Feyzioglu et
al. 1995). The good effect of aid-financed projects (the main micro result) was
counterbalanced, in a number of ‘less open’ economies, by switching of aid into
wasteful consumption expenditures to gratify the preferences of rent-holders.
Fiscal and other management
There
is a risk that aid may erode the tax base, It is a
form of revenue, hence why try to raise money from other sources? indeed Moore (1998) and Brautigam
(2002) demonstrate a negative correlation between aid inflows into a country
and the size of its tax base. Of course, the pattern of causation cannot be
inferred from a mere correlation, and it is clear that there are some African
countries which have escape from the vicious circle, notably
Corruption, governance and state fragility
During
the 1990s, it became clear that the failure of poor, ‘non-convergent’
developing countries to grow or reduce poverty was often associated with the
failure of governments and associated institutions to function in a just and
transparent way. The role of aid broadened from simply the improvement of
efficiency and equity to the improvement of governance, and aid effectiveness
must now be judged not only in terms of its ability, in the short term, to
improve growth performance and reduce poverty, but also in terms of its ability
to improve the performance of governmental and non-governmental institutions
over the long term. Once more this is linked to the concept of conditionality
and opens up the way for not only aid effectiveness to be improved as it is
used in the way donors intend, but from a further knock-on effects from the
improvement in governance.
In
an important paper focused on Africa Brautigam and
Knack (2004) find that in high aid-dependent countries aid provides more
opportunities for, and thereby increases the level of, [the Transparency
International index of] corruption and also reduces tax effort. It thus has a
long-term depressive effect on institutional quality. If these findings are
correct, they imply that what appears now to be a positive short-term effect of
aid on growth and even on poverty is being undermined by long-term forces of
institutional erosion which could cause the short-term effect to be
unsustainable and once more brings into question the whole rationale for aid.
Absorptive
Capacity Problems
Absorptive capacity potentially
affects all developing countries. De Renzio
classifies the constraints under several headings: (i)
Macroeconomic Constraints. Sudden positive increases in aid inflows can cause
an appreciation of the currency, damaging exports, i.e. Dutch disease, (ii)
Institutional and Policy constraints. Aid requires expertise and organisational
infrastructure on the part of the recipient to be able to administer it
effectively. These are often in short supply in many recipient countries and a
sudden increase in aid can stretch them beyond the limit and (iii)Technical and managerial Constraints. Other forms of human
capital are also frequently in short supply. This is evident in for example,
the number of doctors, teachers, nurses, administrators and managers.
Donor behaviour too must share part of the
responsibility for aid’s relative ineffectiveness. In this section we shall
examine this.
Donor Coordination
As
noted by Brautigam and Knack(2004, p. 262) ‘numerous agendas
and projects, with numerous donors, can render ineffective any government’s
efforts to manage its aid resources’. Roodman (2006b)
focuses on problems arising from the recipient’s administrative burden. He
concludes that when projects proliferate beyond a certain point, the effective
marginal utility of aid declines sharply, and can even become negative. There
are also the simple transaction costs of dealing with a large number of donors.
In 2002 there were 25 official bilateral donors, 19 official multilateral
donors and about 350 international NGOs operating in
Donor Consistency
It
is fairly clear from the micro-literature that, other things being equal,
instability in aid flows tends to depreciate its quality. This confronts aid
donors with a fundamental dilemma, since the ‘reactive’ forms of aid have been
introduced, precisely in order to improve the quality of aid by making it
more performance – and need-responsive. In other words there is a
trade-off: if the donor seeks to improve aid quality by reducing volatility,
most ways of doing this – e.g. reducing the penalty for poor performance on
budget support aid – will reduce the donor’s flexibility. An attempt has been
made to finesse this dilemma by Gelb and Eifert
(2006), who propose that, in the event of a negative shock to aid disbursements
caused for example by a fall in the recipient’s CPIA rating, this should be
automatically counteracted by drawings on a buffer fund, which would
stabilise aid disbursements.
Aid volatility has recently come to the fore as an area to analyse. It is argued that if aid is volatile rather than in an even flow this reduces its effectiveness. The following regression table illustrates this (taken from the paper referenced at the end of this section.
Table 7: Regressions of Residuals on Growth
Dependent Variable: Annual Growth of GDP
(Figure 3, equation 1)
|
Estimation method |
OLS |
Random Effects |
Fixed Effects |
Random Effects AR1 |
Fixed Effects AR1 |
RandomEffects IV |
Fixed Effects IV |
|
Constant |
7.092** (5.94) |
10.16** (6.27) |
45.447** (12.43) |
8.362** (4.90) |
38.63 ** (10.12) |
8.794** (4.43) |
36.273** (8.38) |
|
Trend |
-0.0264 (1.57) |
-0.00434 (0.24) |
0.293** (9.33) |
-0.0156 (0.77) |
0.244 (6.76) |
-0.0089 (0.47) |
0.254** (7.84) |
|
Disaster |
-2.772* (2.48) |
-3.550** (3.21) |
-2.564* (2.31) |
-2.419* (2.15) |
-2.107 (1.83) |
-3.829 (3.41) |
-4.741 (4.19) |
|
Aid GDP ratio |
0.0380* (2.38) |
0.0665** (3.29) |
0.164** (5.83) |
0.0664** (3.15) |
0.143 (4.56) |
0.122** (2.93) |
0.433** (4.84) |
|
Inflationt-1 |
-0.00036* (2.45) |
-0.00021 (1.47) |
-0.00016 (1.12) |
-0.00015 (1.00) |
-0.0001 (0.47) |
-0.00025* (1.71) |
-0.00036 (2.39) |
|
Volatility |
-0.101** (4.58) |
-0.105** (4.68) |
-0.148** (5.83) |
-0.0993** (4.50) |
-0.125** (5.20) |
-0.158** (3.80) |
-0.187** (4.49) |
|
Positive Volatility t-1 |
0.0636** (3.24) |
0.0687** (3.55) |
0.0433* (2.26) |
0.0635** (3.34) |
0.0529** (2.75) |
0.0669** (2.67) |
-0.0402 (1.10) |
|
Negative Volatility t-1 |
0.0147 (0.20) |
0.0473 (0.66) |
-0.0127 (0.18) |
0.0382 (0.54) |
0.0120 (1.83) |
0.0704 (1.02) |
0.0414 (0.60) |
|
World Growth
|
0.289** (4.06) |
0.293** (4.26) |
0.209** (3.11) |
0.317** (4.41) |
0.231** (3.20) |
0.303** (4.36) |
0.257** (3.65) |
|
Sub Saharan |
-1.741** (5.59) |
-2.401** (5.22) |
|
-2.140** (4.56) |
|
-2.579** (5.32) |
|
|
|
-1.243** (4.49) |
-1.172** (2.72) |
|
-1.142** (2.63) |
|
-1.163** (2.61) |
|
|
|
-0.883* (1.91) |
0.659 (0.95) |
|
0.826 (1.17) |
|
0.749 (1.04) |
|
|
Log(GDPPCt-2) |
-0.353* (2.26) |
-0.844** (3.82) |
-6.796** (11.93) |
-0.591** (2.59) |
-5.711 (8.48) |
-0.690** (2.65) |
-5.708** (9.21) |
|
F/Wald Observations |
10.40* 2666 |
118.7 2666 |
29.34 2554 |
99.47 2554 |
17.91 2419 |
115.4W 2664 |
30.27 2664 |
**/* denotes significance at the 1% and 5% levels. Wald statistic denoted by W. The figures in parentheses are t statistics. ROUGHLY a value greater than 1.96 indicates that the variable is significant at the 5% level and greater than 2.57 at the 1% level. If the coefficient is positive it indicates that that variable increases growth. If it is negative (like disasters) then an increase in the variable reduces growth. NOTE this is more than correlation we are analysing the causal impoact of aid on growth GIVEN what is happening to all the other variables.
Table 7 presents regression results of the determinants of growth In the first column Aid is significantly positive at the 5% level. However, this is somewhat neutralized by the impact of aid volatility. The results suggest that both positive and negative volatility reduce growth, and in the results we have for the most part combined the two together. However, this does not mean that the two are symmetric in their impact. When we look at lagged volatility there is a substantial difference in their impact. Negative volatility is insignificant, whilst positive volatility is significant and reduces some of the previous harmful impact on aid. If the problems raised by positive volatility are linked with absorptive capacity as is suggested in the literature, these results suggest that these are to an extent short term only and are reversed in subsequent periods. But with negative or downside volatility there is no such reprieve. The disaster variable is also significant in reducing growth. Of the other variables, growth in the OECD countries is transmitted to the developing countries in the current period. There is also a catch-up effect by which growth in developing countries tends to be greater in the poorest countries and the regional dummy variables are also of varying significance. Finally lagged inflation, representing a good policy environment as in the original Burnside and Dollar (2000) paper, is significant with the anticipated negative sign. In the remaining columns we used standard fixed and random effects and then correcting for autocorrelation.
Authors: Hudson, J.; Mosley, P.
Journal: World Development Year: 2008 Volume: 36 Issue:
10 Pages: 2082-2102 Provider: Elsevier
or
http://www.shef.ac.uk/content/1/c6/07/67/88/SERP2007015.pdf
Post Script I
For
many years Sub Saharan Africa has been the region which has posed most
problems, stubbornly stuck in poverty, blessed with huge riches, cursed with corruption,
bad governance and war. Part of the problem is that mineral wealth brings with
it corruption (The Resource Curse). Also these countries are artificial,
created as a result of colonialism. Look at a map of 
http://www.imf.org/external/datamapper/index.php
Post Script II A Slight problem
Climate
change is a real problem. It has many aspects. Most, but not all countries,
will lose. Coastal areas wil ldisappear. Many major cities are on coastal areas
and are at risk. Below is one aspect of this taken from the World Development
Report 2010 “Development and Climate Change”.
Post Script III
CD also look at the use of aid not simply to reduce poverty per se but to promote security which is a fundamental goal in its own right. Although the distinction is not absolute and most low income countries have recent experience of civil conflict. CD report work which shows that the risk of conflict is linked to economic growth, income per capita and dependence upon primary commodity exports. Both aid and policy impact upon these characteristics and hence can be used to reduce the risk of conflict. They report results that (P253) show the effect of a $1 per capita per year additional aid program maintained for 5 years and a one point improvement in good policies also sustained over five years. For the average aid recipient country the conflict risk in a five year period is 11.3%, a one in nine chance that a major civil conflict will be initiated. The aid policy improvement package reduces this risk to 7.9%.and policy reform alone reduces it to 8.6%.
All of this is to illustrate something fundamental. Seldom is there a last word in economics. Seldom is a theory proved entirely correct and that is the end of the matter. The world changes, our resources change, more sophisticated econometric techniques, greater availability of data. An economist writes a paper and if it makes an impact – most do not – then others build upon it, accepting part, rejecting other parts. Hence the CD results on aid and conflict are suggestive and interesting, but just that, not the final word. Theory moves forward in stages, by argument by discussion in this the most challenging and important of academic disciplines.
Table 7: Regressions of
Residuals on Growth
Dependent Variable: Annual
Growth of GDP
(Figure 3, equation 1)
|
Estimation
method |
OLS |
Random
Effects |
Fixed
Effects |
Random
Effects AR1 |
Fixed Effects AR1 |
RandomEffects IV |
Fixed Effects IV |
|
Constant |
7.092** (5.94) |
10.16** (6.27) |
45.447** (12.43) |
8.362** (4.90) |
38.63 ** (10.12) |
8.794** (4.43) |
36.273** (8.38) |
|
Trend |
-0.0264 (1.57) |
-0.00434 (0.24) |
0.293** (9.33) |
-0.0156 (0.77) |
0.244 (6.76) |
-0.0089 (0.47) |
0.254** (7.84) |
|
Disaster |
-2.772* (2.48) |
-3.550** (3.21) |
-2.564* (2.31) |
-2.419* (2.15) |
-2.107 (1.83) |
-3.829 (3.41) |
-4.741 (4.19) |
|
Aid GDP
ratio |
0.0380* (2.38) |
0.0665** (3.29) |
0.164** (5.83) |
0.0664** (3.15) |
0.143 (4.56) |
0.122** (2.93) |
0.433** (4.84) |
|
Inflationt-1 |
-0.00036* (2.45) |
-0.00021 (1.47) |
-0.00016 (1.12) |
-0.00015 (1.00) |
-0.0001 (0.47) |
-0.00025* (1.71) |
-0.00036 (2.39) |
|
Volatility |
-0.101** (4.58) |
-0.105** (4.68) |
-0.148** (5.83) |
-0.0993** (4.50) |
-0.125** (5.20) |
-0.158** (3.80) |
-0.187** (4.49) |
|
Positive
Volatility t-1 |
0.0636** (3.24) |
0.0687** (3.55) |
0.0433* (2.26) |
0.0635** (3.34) |
0.0529** (2.75) |
0.0669** (2.67) |
-0.0402 (1.10) |
|
Negative
Volatility t-1 |
0.0147 (0.20) |
0.0473 (0.66) |
-0.0127 (0.18) |
0.0382 (0.54) |
0.0120 (1.83) |
0.0704 (1.02) |
0.0414 (0.60) |
|
World Growth |
0.289** (4.06) |
0.293** (4.26) |
0.209** (3.11) |
0.317** (4.41) |
0.231** (3.20) |
0.303** (4.36) |
0.257** (3.65) |
|
Sub
Saharan |
-1.741** (5.59) |
-2.401** (5.22) |
|
-2.140** (4.56) |
|
-2.579** (5.32) |
|
|
|
-1.243** (4.49) |
-1.172** (2.72) |
|
-1.142** (2.63) |
|
-1.163** (2.61) |
|
|
|
-0.883* (1.91) |
0.659 (0.95) |
|
0.826 (1.17) |
|
0.749 (1.04) |
|
|
Log(GDPPCt-2) |
-0.353* (2.26) |
-0.844** (3.82) |
-6.796** (11.93) |
-0.591** (2.59) |
-5.711 (8.48) |
-0.690** (2.65) |
-5.708** (9.21) |
|
F/Wald Observations |
10.40* 2666 |
118.7 2666 |
29.34 2554 |
99.47 2554 |
17.91 2419 |
115.4W 2664 |
30.27 2664 |
**/* denotes significance at the 1% and 5%
levels. Wald statistic denoted by W. The figures in parentheses are
t statistics. ROUGHLY a value greater than 1.96 indicates that the variable is
significant at the 5% level and greater than 2.57 at the 1% level. If the coefficient is positive it indicates
that that variable increases growth. If it is negative (like disasters) then an
increase in the variable reduces growth. NOTE this is more than correlation we
are analysing the causal impoact of aid on growth
GIVEN what is happening to all the other variables.
