The Economic and Social Review, Vol. 20, No. 2, January 1989, pp. 165-188
Price Determination in Ireland: Effects of Changes
in Exchange Rates and Exchange Rate Regimes*
T I M C A L L A N and J O H N F I T Z G E R A L D
The Economic and Social Research Institute, Dublin
Abstract: This paper finds that the wholesale price of Irish manufacturing o u t p u t is closely tied t o
foreign prices i n the long r u n ; domestic wage costs generally do not exert a significant influence. I n the short run changes i n foreign currency prices affect Irish prices more rapidly than do changes i n exchange rates. There is evidence of a change i n pricing behaviour on Ireland's entry i n t o the EMS. German out put prices are now as i m p o r t a n t as U K prices i n determining Irish prices.
I I N T R O D U C T I O N
I
t is w e l l established b y n o w that short-run purchasing p o w e r p a r i t y does n o t h o l d between major countries (see, for example, Frenkel, 1981 or Isard, 1977). This means that one cannot define an empirical counterpart t o the single w o r l d price o f simple small open economy (SOE) models. Several approaches can be adopted t o dealing w i t h this p r o b l e m o f m o d e l l i n g the i n f l u ence o f foreign prices o n the domestic prices o f a small c o u n t r y . When Ireland's exchange rate was f i x e d at p a r i t y w i t h sterling the d o m i n a n t approach adopted i n empirical w o r k was the use o f prices i n the U K , Ireland's largest trade partner, t o p r o x y w o r l d prices. The alternative, o f using a trade-weighted basket o f the prices o f Ireland's major trade partners, has also been given con siderable a t t e n t i o n .I n this paper we reassess these results, a l l o w i n g for differences i n the short-r u n short-response t o exchange short-rates and foshort-reign pshort-rices, and foshort-r possible changes
i n the relative i m p o r t a n c e o f U K and German prices associated w i t h Ireland's membership o f the European M o n e t a r y System (EMS). We use quarterly data on the wholesale prices o f the o u t p u t o f m a n u f a c t u r i n g i n d u s t r y . Wholesale prices do n o t include V A T or retail d i s t r i b u t i o n costs, so they p r o v i d e more direct evidence o n the influence o f foreign prices than consumer prices, w h i c h also include a m u c h larger non-traded c o m p o n e n t . We allow for the possibility o f differential response t o foreign currency price changes and exchange rate changes. We investigate possible changes i n the relative importance o f dif ferent countries, or other changes i n the influence o f foreign prices arising f r o m or associated w i t h Ireland's EMS membership. Specifically, we examine econometric tests o f the f o l l o w i n g questions:
(1) A r e Irish prices p r o p o r t i o n a l t o foreign prices i n Irish currency i n the l o n g run?
(2) D o domestic u n i t wage costs exert at least a short-run influence o n prices, as a mark-up m o d e l o f p r i c i n g w o u l d suggest?
(3) D o Irish prices react similarly i n the short r u n t o changes i n foreign currency prices and t o exchange rate changes, or is there a differential response?
(4) Is there evidence o f an asymmetric response o f Irish prices t o increases and decreases i n foreign prices?
(5) W h i c h specification o f foreign price variables performs best: U K prices alone, a trade-weighted index, or the prices o f several major com petitors?
(6) Has e n t r y t o the EMS l e d t o a change i n the relative influence o f dif ferent countries, or t o any other changes i n the relationship between foreign and domestic prices?
The remainder o f the paper is stuctured as follows. A brief review o f pre vious w o r k o n price d e t e r m i n a t i o n i n I r e l a n d is given i n Section I I . Section I I I describes the basic m o d e l and Section I V gives results on the six m a i n ques tions o u t l i n e d above, based o n analysis o f wholesale price indices for manu facturing i n d u s t r y treated as an aggregate. Section V deals briefly w i t h the results o f a p p l y i n g the basic m o d e l t o the data o n the o u t p u t prices o f i n d i vidual i n d u s t r i a l sectors. Section V I draws together the m a i n conclusions.
I I R E V I E W O F P R E V I O U S W O R K
The conclusions f r o m this w o r k , as summarised b y H o n o h a n and F l y n n ( 1 9 8 6 ) , were t h a t " I r i s h price i n f l a t i o n was determined b y , and more or less equal t o , U K i n f l a t i o n . " (p. 175).
This evidence l e d t o a widespread expectation t h a t EMS membership w o u l d r a p i d l y result i n the Irish rate o f i n f l a t i o n being determined i n a similar fashion by German i n f l a t i o n . The experience o f EMS membership has been rather different. Whereas, i n the sterling l i n k p e r i o d , there was n o u n c e r t a i n t y con cerning the future value o f the sterling exchange rate, the relationship between the I r i s h p o u n d and the Deutschmark has n o t remained f i x e d over t i m e and firms have been faced w i t h u n c e r t a i n t y about future rates o f exchange between the Irish p o u n d and other EMS currencies. The u n c e r t a i n t y concerning the rate o f exchange between the Irish p o u n d and sterling, w h i c h has remained outside the EMS, has been even greater. The p a t t e r n o f I r i s h i n f l a t i o n since j o i n i n g the EMS cannot be explained b y the simple SOE m o d e l o f price deter m i n a t i o n : i t is no longer closely t i e d t o the U K rates, and i t has n o t m i r r o r e d German rates. A t first sight, this suggested a change i n the foreign influence on I r i s h prices. However, H o n o h a n and F l y n n ' s analysis o f the 1972-1984 p e r i o d showed that a more flexible dynamic specification o f the influence of foreign prices, using a trade-weighted index rather than s i m p l y U K prices, c o u l d y i e l d an adequate m o d e l for b o t h the sterling l i n k and EMS sub-periods.
Neither H o n o h a n and F l y n n ' s trade-weighted approach, nor the earlier w o r k based o n U K prices, examined the question o f whether the relative i m p o r t a n c e o f different countries' prices i n determining Irish prices has changed over t i m e , i n particular i n response t o Ireland's membership o f the European M o n e t a r y System (EMS). The use o f trade weights, w h i l e valuable for m a n y purposes, implies that the relative i m p o r t a n c e changes i f and o n l y i f trade patterns change, thus s i m p l y imposing an answer t o the question. The use o f U K prices does n o t even allow this l i m i t e d a m o u n t o f f l e x i b i l i t y .
F i t z G e r a l d (1983) e x p l o r e d the issue using data disaggregating foreign price influence b y c o u n t r y . However, that analysis was l i m i t e d b y the small number o f post-EMS observations. The approach adopted there, and e x p l o r e d i n more detail here, was t o regress Irish o u t p u t prices against the prices o f Ireland's major c o m p e t i t o r s , n o t a b l y the U K , Germany and the U S A , a l l o w i n g the relative importance o f different countries t o be more d i r e c t l y determined f r o m the data. A priori one w o u l d expect considerable difference i n behaviour between domestic Irish firms and foreign-owned firms, and also between dif ferent sectors. I n order t o examine this p o s s i b i l i t y , we repeat the analysis for selected sectors o f m a n u f a c t u r i n g i n d u s t r y (as d i d F i t z G e r a l d ) , some o f w h i c h are d o m i n a t e d b y m u l t i n a t i o n a l s , and others b y domestic f i r m s .1
FitzGerald also f o u n d a stronger response t o sterling price changes i n the p e r i o d 1979-81 than t o sterling exchange rate changes, although attempts t o disaggregate the exchange rate and foreign currency price changes for other countries gave less satisfactory results. This analysis strongly suggested further investigation o f differential effects o f exchange rate and foreign currency price changes, a topic neglected i n other analyses o f Irish price d e t e r m i n a t i o n .
Leddin's ( 1 9 8 8 ) more recent analysis has investigated some o f the issues addressed here, also using wholesale price index data for manufacturing i n d u s t r y . However, Leddin's specification imposed some crucial r e s t r i c t i o n s2
w h i c h our tests indicate are rejected by the data. His conclusion, that short-r u n pushort-rchasing poweshort-r p a short-r i t y does n o t h o l d , is supposhort-rted by oushort-r moshort-re extensive tests, b u t our results show t h a t a long-run version o f purchasing power p a r i t y ( n o t tested b y L e d d i n ) may h o l d .
I l l B A S I C M O D E L A N D T H E D A T A
We begin w i t h t w o alternative general models w h i c h express the long-run d e t e r m i n a t i o n o f the level o f prices i n I r e l a n d ; these t w o models represent the polar cases o f the pure SOE m o d e l and the closed economy alternative where prices are determined as a mark-up o n i n p u t costs.
pi * R = fi (piIV >i = 1>n) W
PI*R= f2( Wj; j = l , m ) (2)
where
P*R = the long-run desired wholesale price for Irish manufacturing o u t p u t
P. = the wholesale price for manufacturing o f c o u n t r y i R. = the rate o f exchange between Ireland and c o u n t r y i Wj = the price index for i n p u t factor j .3
I f prices are p u r e l y externally determined, as i n E q u a t i o n ( 1 ) , then, assum ing that purchasing power p a r i t y holds i n the long r u n , Irish prices w i l l be some weighted average o f the prices o f a possible range o f other countries, all expressed i n Irish p o u n d terms. I f , o n the other hand, they are purely deter m i n e d as a mark-up o n domestic costs, then E q u a t i o n (2) is appropriate. I n examining the behaviour o f o u t p u t prices we test for whether either of these t w o long-run models o f price d e t e r m i n a t i o n o n their o w n or, alternatively,
2. These include, for example, the imposition of identical coefficients o n foreign currency prices and exchange rates.
some m i x t u r e o f the t w o is appropriate. This is done b y c o m b i n i n g Equations (1) and (2) i n t o the composite E q u a t i o n ( 3 ) .
P*R = f3( P . Ri >W . ; i = l , n a n d j = l , m ) (3)
T a k i n g E q u a t i o n (3) as the u n d e r l y i n g long-term relationship determining Irish o u t p u t prices, we a l l o w for differential short-run effects f r o m the prices and exchange rates o f the U K , Germany (GR) and the US, and f r o m the price of domestic factor i n p u t s . This is i m p l e m e n t e d using an error-correction type t e r m w h i c h allows a long-run influence f r o m prices i n each o f these countries and, possibly, f r o m domestic costs; the o n l y r e s t r i c t i o n imposed at this stage was that o f long-run homogeneity, tested later. The structure o f the m o d e l is, therefore, similar t o that estimated on trade-weighted data b y H o n o h a n and F l y n n ( 1 9 8 6 ) . The m o d e l can be w r i t t e n as follows where L stands for the lag operator, PJ R for the Irish price, for the price o f c o u n t r y j , R. for
the bilateral exchange rate w i t h respect t o c o u n t r y j , W for u n i t wage costs, and a, |3;J, y.^, e.^ r?j and X are all parameters t o be estimated:
A l o g ( PI R) = \ £ PijL1 { A l o g ( P ) } + I I y L M A l o ^ R ) }
i= 0 j = l J J 1= 0 j = l J J
The /3.j, 7 j j , and e; terms give the short-run responses t o changes i n foreign
currency prices and exchange rates i n the three countries and t o domestic u n i t wage costs, w h i l e the last t e r m is an error-correction mechanism ( E C M ) b y w h i c h Irish prices are b r o u g h t i n t o a long-run e q u i l i b r i u m r a t i o w i t h a weighted average4 o f the Irish currency prices o f the three countries and
labour costs. The constant i n the E C M t e r m allows for this e q u i l i b r i u m r a t i o t o be some constant otner than u n i t y ; i n the estimated equations we r e p o r t the coefficient for a. This m o d e l allows for deviations f r o m purchas ing p o w e r p a r i t y i n the short r u n . H o m o g e n e i t y i n the l o n g r u n is tested i n a later section b y adding an unrestricted t e r m i n L4 ( l o g ( PI R) ) . A p p r o p r i a t e
seasonal dummies were i n c l u d e d i n all the equations.
The data o n the wholesale price o f the different categories o f m a n u f a c t u r i n g i n d u s t r y o u t p u t , i n c l u d i n g t o t a l o u t p u t , are taken f r o m the CSO databank E O L A S . The data o n u n i t wage costs for each sector are derived f r o m a similar source. The data o n exchange rates are p a r t l y taken f r o m the E O L A S data bank and p a r t l y f r o m the D e p a r t m e n t o f Finance databank o f series f r o m the O E C D Main Economic Indicators p u b l i c a t i o n . The o u t p u t prices for each o f the industrial sectors i n the other countries are taken f r o m D e p a r t m e n t o f Finance databank o f series f r o m the O E C D Quarterly Indicators of Industrial Activity p u b l i c a t i o n . The exchange rate data for the three m a i n countries (the U S A , the U K , and West Germany) are p e r i o d averages whereas the exchange rate data for certain other countries t r i e d i n the disaggregated equations are averages o f end m o n t h l y values.5 The availability o f the wholesale price indices
basically determined the data p e r i o d as the first quarter o f 1975 t o the t h i r d quarter o f 1987. The estimation p e r i o d , unless otherwise indicated, is the first quarter o f 1976 t o the t h i r d quarter o f 1987.
I V R E S U L T S F O R T O T A L M A N U F A C T U R I N G I N D U S T R Y 4.1 Preliminary Tests of Co-integration
Before estimating the m o d e l ( E q u a t i o n (4)) we check that the t i m e series properties o f the data satisfy the constraints w h i c h Engle and Granger (1987) have shown are i m p l i e d b y the m o d e l . The basic procedure suggested is i n t w o steps. The first step is t o test that the i n d i v i d u a l series share a c o m m o n order o f i n t e g r a t i o n (i.e., are each stationary after differencing the same number o f t i m e s ) ; the second step is t o test that the series move together i n the l o n g r u n (i.e., are " c o - i n t e g r a t e d "6) .
Since the models u l t i m a t e l y used i n this section involve j u s t U K and German prices as explanatory variables, they i m p l y a constant long-run relationship between I r i s h prices and a weighted average o f U K and German prices, o n a c o m m o n currency basis. Therefore, we examine first o f all the order o f i n t e g r a t i o n for I r i s h , U K and German prices, all i n Irish p o u n d terms. We test the hypothesis that the series is a r a n d o m w a l k , using the test statistics derived by D i c k e y and Fuller ( 1 9 7 9 ) . The hypothesis that the series is a r a n d o m w a l k w i t h o u t d r i f t (so that the series is "integrated o f order 1") is tested b y $2,
w h i l e 4>3 tests the hypothesis that the first differenced series is a r a n d o m w a l k
w i t h d r i f t .
The test statistics for the f u l l p e r i o d , and for the post-EMS p e r i o d are shown i n Table 1.
5. The data available in the databank.
T a b l e 1: Augmented Dickey-Fuller Test Statistics for Hypothesis that Irish Pound Price
Series are Integrated of Order 1
Full Period EMS Period
(45 obs.) (33 obs.)
* 2 * 3 1 2 * 3
L ° g ( PT R) 2.45 2.55 1.40 1.34
L° g (PU K X RU K )
2 . 7 8 2 . 5 8 4 . 4 7 5 . 5 5
L o g ( PG R X RGR) 3 . 7 2 0 . 9 2 3 . 9 8 3 . 1 6
C r i t i c a l values f r o m D i c k e y a n d F u l l e r ( 1 9 8 1 ) :
50 obs. 25 obs.
5% 5 . 1 3 6.73 5 . 6 8 7 . 2 4
10% 4.31 5.61 4 . 6 7 5.91
These test statistics do n o t reject the hypotheses t h a t each o f the Irish p o u n d price series is integrated o f order 1, even at the 10 per cent level. H o w ever, the results m a y be sensitive t o the p e r i o d i c i t y o f the data, or perhaps t o the exact sub-period chosen, as the results o f T h o r n elsewhere i n this Review m i g h t suggest. The statistics also appear t o be q u i t e sensitive t o certain choices w h i c h m u s t be made w i t h i n the test procedure, an issue t o w h i c h we n o w t u r n . The augmented D i c k e y - F u l l e r regression f r o m w h i c h these statistics are derived involve, for the l o g ( PI R) series, lagged terms i n A l o g ( PI R) . Granger
(1986) indicates that the order o f the autoregressive terms should be large enough t h a t the residuals f r o m the augmented Dickey-Fuller regression are empirical w h i t e noise. The best procedure for establishing the order o f this autoregression has y e t to be established. The procedure adopted here was t o start w i t h four lags, because the quarterly price data are k n o w n t o involve seasonal factors, and add further lags i f the F-form o f the Lagrange M u l t i p l i e r test indicates a u t o c o r r e l a t i o n o f the residuals at the 5 per cent level. I n all cases where a d d i t i o n a l lags were i n t r o d u c e d , they had h i g h l y significant t-statistics, and the F-testno longer i n d i c a t e d the presence o f a u t o c o r r e l a t i o n . I f the order of the autoregression was instead set and m a i n t a i n e d at four, the hypothesis that the Irish p o u n d price o f German goods is integrated o f order 1 w o u l d be rejected for the full p e r i o d , t h o u g h n o t for the post-EMS p e r i o d .
bivariate relationships, using a simple "co-integrating regression" w h i c h regresses one variable o n the other plus a constant t e r m . Some test statistics for higher order cases have been p r o d u c e d b y Engle and Y o o ( 1 9 8 7 ) , using a multivariate co-integrating regression.
We examine first of all whether I r i s h prices share a c o m m o n t r e n d or long-r u n e q u i l i b long-r i u m long-relationship w i t h eithelong-r U K olong-r Gelong-rman plong-rices ( i n Ilong-rish p o u n d terms), and then whether there is a c o m m o n long-run e q u i l i b r i u m relationship between Irish prices and some weighted average o f U K and German prices. The test statistic is based o n a regression o f the first differences o f the residuals f r o m the co-integrating regression on autoregressive terms, together w i t h the lagged value of the levels of these residuals (again, see T h o r n , this Review, or Engle and Granger, 1987, for details). The choice o f the number o f auto regressive terms t o be i n c l u d e d i n this regression raises issues parallel t o those discussed above, and the sensitivity o f the results is similar. The hypothesis of co-integration is rejected i f the calculated t-statistic for the lagged level o f the residuals does n o t exceed the critical values. The d i s t r i b u t i o n o f this A u g m e n t e d Dickey-Fuller ( A D F ) test statistic is n o t i n fact given b y the t - d i s t r i b u t i o n , b u t critical values are tabulated b y Engle and Y o o ( 1 9 8 7 ) . Table 2 gives the results for the f u l l p e r i o d , and the post-EMS p e r i o d .
T a b l e 2: Tests of Co-integration between log(Pffj), log(Pnf, X RUK), and log(PGR X RGR)
Full Period EMS Period
(44 obs.) (27 obs.)
ADF ADF
PIRa n d PUKX R UK 1.82 1.29
PIRa n d PGRX RGR 1.47 3 . 1 6
Critical values^ 50 obs. 25 obs.
5% 3 . 2 9 N A
10% 2.90 N A
PI R 'PU K X RU K 'a n d PG R X RG R 3.42
3 . 5 0
Critical values 50 obs. 2 5 obs.
5% 3.75 N A
10% 3 . 3 6 N A
The results for the f u l l p e r i o d illustrate the possibility that Irish prices m a y n o t be co-integrated w i t h either U K prices or German prices, b u t c o u l d be co-integrated w i t h a weighted average o f the t w o : the o n l y statistic w h i c h is significantly different f r o m zero (thereby i n d i c a t i n g co-integration) is that based o n the hypothesis o f a long-run e q u i l i b r i u m relationship between I r i s h prices and both U K and German prices. While critical values are n o t y e t avail able for the n u m b e r o f observations corresponding t o the post-EMS p e r i o d , the results suggest a stronger role for German prices, b u t w o u l d still a l l o w for a long-run l i n k t o some weighted average o f U K and German prices.
Given t h a t there is support f r o m this co-integration analysis for the existence of a l o n g - r u n e q u i l i b r i u m relationship between Irish prices and a weighted aver age o f U K and German prices, we proceed t o the estimation o f E q u a t i o n ( 4 ) .8
4.2 Basic Model: Dynamic Specification and Stability Tests
E q u a t i o n (4) was applied t o the price o f o u t p u t o f t o t a l m a n u f a c t u r i n g . The u n i t wage cost variable p r o v e d insignificant. This suggests that the long-r u n long-relationship d e t e long-r m i n i n g Ilong-rish plong-rices is closelong-r t o the pulong-re SOE m o d e l , described b y E q u a t i o n ( 1 ) , than t o the m i x e d m o d e l u n d e r l y i n g E q u a t i o n ( 3 ) .9
Data for prices and exchange rates o f three different countries, the U K , Germany and the U S A , were t r i e d i n the equation. However, i n estimating E q u a t i o n ( 4 ) , i t was apparent that neither US prices n o r the dollar exchange rate p l a y e d any significant role i n the d e t e r m i n a t i o n o f wholesale prices at an aggregate level. This is somewhat surprising, given the fact that price-setting i n dollars has become increasingly i m p o r t a n t , especially i n the office machines and data processing equipment and pharmaceuticals sectors, w h i c h have g r o w n r a p i d l y over the p e r i o d t o f o r m a significant part o f m a n u f a c t u r i n g i n d u s t r y . However, there are inherent difficulties i n measuring the prices o f these sectors. I n particular, the problems posed b y r a p i d q u a l i t y change i n the c o m p u t i n g i n d u s t r y have l e d t o the exclusion o f its prices f r o m the whole sale price index i n I r e l a n d , so that the lack o f US influence m a y reflect the exclusion f r o m the index o f the sector where prices are most obviously set i n US dollars.
Three lags o n the change i n German and U K prices and exchange rates were i n i t i a l l y i n c l u d e d , together w i t h an error c o r r e c t i o n t e r m lagged four periods and seasonal dummies Sf c. These i n i t i a l estimates are r e p o r t e d as E q u a t i o n ( 5 ) .1 0
8. We do not perform similar co-integration analysis of the other variables i n that equation (US prices, wages) since the simpler tests used i n the next section indicate that they should be discarded.
9. As Baker (1985) has pointed out, the rapid g r o w t h of high technology sectors o f the economy in the 1980s poses problems for the use o f an aggregate wage cost variable covering all o f manufacturing industry. As discussed i n the next section, when examined on a disaggregated basis, wage costs proved to be a significant factor determining prices i n the long r u n in t w o sectors, i.e., electrical engineering and paper and paper products.
A 1° g (PlR) = . ViGRLi^log (PG R ) >+ . V ^ L ^ A l o g l R ^ ) }
1= 0 1= u
+ . 1 PI U KLi{ A l o g ( Pl K) } + _i yiVKV{Mog(RVK)}
+ a+
I
fjS. (5)i = 0
The f i t o f the relationship is very good, and most coefficients have significant values o f the expected (i.e., positive) sign; the three coefficients w i t h negative signs are n o t significantly different f r o m zero. The long-run coefficients are p a r t i c u l a r l y well-defined.
Equation 5 Results11
Coefficient: i=0 i = l i=2 i=3
a 1.379
( 4 . 7 )
0 i G R 0 . 5 7 5 0 . 0 9 2 - 0 . 2 8 2 0 . 1 2 2
0 i G R
(2.1) (0.3) ( - 0 . 8 ) (0.4)
T i G R 0 . 1 6 5 0.171 0 . 0 9 6 0 . 0 8 4
T i G R
(4.6) (4-0) (2.3) (1.8)
0 i U K 0 . 1 6 9 - 0 . 1 1 7 0 . 1 9 3 - 0 . 0 6 3
0 i U K
(1.1) ( - 0 . 7 ) (1.1) ( - 0 . 5 )
I ' i U K 0 . 0 4 8 0.131 0 . 0 7 9 0 . 0 8 9
I ' i U K
( 1 . 2 ) (3.4) (2.1) (2.6)
0G R
0 . 1 3 5
( 4 . 3 )
^ U K 0 . 0 9 6
(4.0)
0 . 0 1 8
(5-3)
0 . 0 1 2
(2.8)
0 . 0 0 4 (1.1)
R2 = . 9 4 0 S E R = . 0 0 5 3 4 R2 = .887
M e a n = . 0 2 2 9 S . D . = .0 1 7 1 C h o w F [ 1 2 , 3 5 ] = 1 . 2
A R F [ 4 , 2 0 ] = . 9 0 6 D F F I T S = 2 . 9 2 Het F [ l , 4 4 ] = .6 8 4
The diagnostic statistics are generally satisfactory. The F-versions o f the Lagrange M u l t i p l i e r tests do n o t indicate that the errors are autocorrelated or heteroscedastic. The relationship is relatively robust t o the d r o p p i n g o f the last 12 observations, as i n d i c a t e d b y the C h o w test: m o r e detailed analysis o f possible structural breaks related t o the EMS regime is taken up later i n this section. The D F F I T S statistic is above the critical level suggested b y Krasker, K u h and Welsch (1983) i n d i c a t i n g t h a t at least one observation is exerting significant leverage o n the results. However, w h e n estimated using b o u n d e d influence regression, the results were l i t t l e changed (Krasker, K u h and Welsch,
1983).
The strength o f the influence f r o m Germany, and the weakness o f the U K influence are interesting. The coefficients o n the E C M terms indicate that the long-run weight o n German prices i n Irish p o u n d terms is 0.6 while that on the U K is 0.4. O n the basis o f trade shares one m i g h t have expected the U K influence t o be g r e a t e r .1 2 The difference between the estimated
short-r u n and long-short-run effects i n this and otheshort-r equations accounts foshort-r the auto c o r r e l a t i o n f o u n d b y L e d d i n (1988) i n a specification w h i c h d i d n o t a l l o w for these differential dynamics.
We w a n t t o investigate b o t h the dynamic specification o f this relationship, and its s t a b i l i t y across the change o f exchange rate regime. T o make our results independent o f the order i n w h i c h we u n d e r t o o k these tasks, we have p e r f o r m e d F-tests for a series o f nested models. T h e y involve simplifications of the dynamic structure and alternative changes associated w i t h EMS e n t r y . I n Figure 1 the dynamic structure o f E q u a t i o n ( 5 ) , w i t h three lags o n all short-run coefficients, is labelled A ; the s i m p l i f i c a t i o n (suggested b y the t-statistics) w h i c h drops the three lags o n the price terms is labelled B ; a corresponding s i m p l i f i c a t i o n , d r o p p i n g the lags o n the exchange rate terms is labelled Z ; w h i l e the final s i m p l i f i c a t i o n , d r o p p i n g the t h i r d lag on the German exchange rate, is labelled C. A specificiation w h i c h allows b o t h long- and short-run relationships t o change after EMS e n t r y is labelled 1 ;1 3 a specifica
t i o n w h i c h allows o n l y the long-run relationship t o change, as shown i n E q u a t i o n (6) below, is labelled 2; and the specification w h i c h imposes the same relationship over the f u l l p e r i o d (as i n E q u a t i o n (5)) is labelled 3. Thus we have a 4 X 3 m a t r i x o f possible specifications, w i t h E q u a t i o n (5) repre sented b y A 3 . Figure 1 b e l o w indicates the specifications preferred b y the F-test at the 5 per cent level: the calculated statistics are given i n A p p e n d i x 1.
12. This issue is discussed later i n Section 4.6.
F i g u r e 1: Summary of Results of Nested Tests
(Significance level: 5%)
(1) (2) (3)
Short- & Long- Long-run Short- & Long-run Relationships Relationship run Relationships
Changed by EMS Changed by EMS Unchanged by EMS
Z . D r o p lags o n R Z l => Z 2 «= Z 3
A . 3 lags o n P a n d R A l A 2 <= A 3
*
B . D r o p lags o n P B l B 2 => B 3
C . D r o p 3 r d lag
»
1to n R G R C I C 2 C 3
These F-tests c o n t a i n several features o f interest. The d r o p p i n g o f the lags on the price terms is never rejected; corresponding tests o n d r o p p i n g the lags on the exchange rate terms are always strongly rejected. This is strong evidence for an a s y m m e t r y between the effects o f foreign price changes and exchange rate changes o n Irish wholesale prices: the exchange rate effects are more spread o u t over t i m e , w h i l e the price effects are concentrated i n the current p e r i o d . F u r t h e r testing o n this issue is carried o u t i n Section 4.4 b e l o w .
Tests for a s t r u c t u r a l break after e n t r y t o the EMS are accepted for all b u t one specification at the 5 per cent level, and for all at the 10 per cent level. The hypothesis that this structural break is concentrated on the long-run coefficients cannot be rejected at the 5 per cent level for any o f the specifi cations considered. This strongly suggests that there was a change i n the d e t e r m i n a t i o n o f Irish wholesale prices subsequent t o EMS e n t r y . The fact that the short-run coefficients do n o t change significantly may reflect the d i f f i c u l t y i n i d e n t i f y i n g the exact short-run relationships, whereas the longer-r u n longer-relationships diffelonger-r molonger-re clealonger-rly.
N o m a t t e r w h a t the starting p o i n t i n the m a t r i x o f specifications o f Figure 1, f o l l o w i n g the results o f the nested tests leads t o either specification C2 or B3. This a m b i g u i t y can be resolved b y m o v i n g t o a 10 per cent significance level for the tests. The o n l y change i n terms o f Figure 1 is that the arrow f r o m B2 t o B3 is reversed. This means that n o matter w h a t the starting p o i n t ,1 4
the arrows n o w p o i n t eventually t o specification C2, w h i c h is r e p o r t e d b e l o w as E q u a t i o n ( 6 ) .
I n the preferred specification, E q u a t i o n ( 6 ) , the long-run relationship is a l l o w e d t o change o n j o i n i n g the E M S . This is i m p l e m e n t e d b y m u l t i p l y i n g a d u m m y1 5 b y the error c o r r e c t i o n terms for the U K and Germany.
A l o g ( PI R) = i =l Pi G RLi{ A l o g ( PG R) } + T i G RLi{ A l o g ( RG R) }
+ .}Q 0 i U KL i{A 1° g (PU K ) }+ j * , 7 i U KLi{Alog ( RUK) }
^ C R ^ o g Q^*) + 9 O T L * l o g ( ^ « )
/ + g C R L 4 l o g ( P G R ^ R G R ) x D E M S
+ 6U KL4l o g (?UK p RUK) X D m s + a + I f . S; (6)
While the t-statistic o n the coefficients o f the t w o " s t r u c t u r a l b r e a k " vari ables are i n d i v i d u a l l y significant o n l y at the 10 per cent level, the F-test o f the j o i n t hypothesis t h a t b o t h are equal t o zero yields a value o f 5.59, w h i c h is significant at the 1 per cent level. As shown i n Table 3, these results i m p l y t h a t the long-run weights o n German and U K prices changed o n j o i n i n g the EMS as the importance o f U K o u t p u t prices i n d e t e r m i n i n g I r i s h o u t p u t prices was greatly reduced and there was a corresponding increase i n the importance o f German o u t p u t prices.
A break i n this estimated relationship, i n v o l v i n g f i x e d weights o n each c o u n t r y ' s price indices has t w o possible explanations. I f the u n d e r l y i n g w o r l d price d e t e r m i n i n g Irish o u t p u t prices were, a trade-weighted average o f prices i n our m a i n t r a d i n g partners the significance o f the dummies c o u l d merely reflect a change i n trade weights, i.e., the higher weight o n German prices i n the later p e r i o d c o u l d be due solely t o its increasing i m p o r t a n c e i n I r i s h trade. This e x p l a n a t i o n can, however, be discounted, because o f the evidence f r o m the trade-weighted regressions below.
The alternative e x p l a n a t i o n , supported b y these results, is t h a t there was a change i n price setting behaviour after Irish e n t r y i n t o the EMS. This suggests that the relative i m p o r t a n c e o f German prices as against U K prices i n Irish f i r m s ' p r i c i n g decisions may have been altered f o r more i n s t i t u t i o n a l reasons, perhaps associated w i t h a reversal i n the relative s t a b i l i t y o f the Deutschmark and Sterling n o m i n a l exchange rates under the n e w regime.
Equation 6 Results
Coefficient: i=0 i = l i=2 i=3
a 1.050
( 3 . 7 )
0 . 5 0 0 ( 2 . 8 )
^ i G R 0 . 1 1 8
( 3 . 6 )
0 . 1 1 4
( 3 . 0 )
0 . 0 6 1
( 1 . 8 )
0 i U K 0 . 2 7 0
( 2 . 5 )
I ' i U K 0 . 0 6 7
( 2 . 5 )
0 . 1 3 5
( 5 . 1 )
0 . 0 8 1 0.081 ( 3 . 0 ) ( 3 . 3 )
9G R
0 . 0 2 4
( 0 . 6 )
9U K
0 . 1 6 7
( 2 . 9 )
5G R
0 . 0 6 1 ( 1 . 5 )
5U K
- 0 . 0 7
( - 1 . 4 )
? i 0 . 0 1 6 0.011 0 . 0 0 3
( 5 . 8 ) ( 4 . 1 ) ( 1 . 3 )
R2 = . 9 4 2 S E R = . 0 0 4 9 0 R2 = . 9 1 0
A R F [ 4 , 2 5 ] = 1.36 D F F I T S = 1.99 H e t F [ l , 4 4 ] = 1.12
T a b l e 3 : Weights on German and UK Prices
Germany UK
P r e - E M S 0.13 0.87
P o s t - E M S 0.47 0 . 5 3
4.3 Tests of Long-run Homogeneity
The tests o f co-integration i n Section 4 . 1 have i n d i c a t e d support for a long-r u n e q u i l i b long-r i u m long-relationship between Ilong-rish plong-rices and a weighted avelong-rage o f the Irish p o u n d prices o f German and U K manufactures. This is a necessary, b u t insufficient, c o n d i t i o n for long-run homogeneity t o h o l d . The r e s t r i c t i o n o f long-run homogeneity, or relative purchasing p o w e r parity, has been imposed i n the models o f the previous section: the long-run s o l u t i o n o f these models involves a constant r a t i o between Irish prices and a (geometric) weighted average o f U K and German prices i n Irish p o u n d terms. Given the nature o f price indices i t is n o t possible t o draw any conclusions concerning absolute purchasing p o w e r p a r i t y f r o m these results.
A stronger test o f the homogeneity r e s t r i c t i o n can be derived b y i n c l u d i n g as an a d d i t i o n a l variable the f o u r t h lag o f the level o f Irish prices: this allows the long-run relationship t o be a non-constant r a t i o . I f the r e s t r i c t i o n is rejected by this test i t w o u l d suggest the omission o f some i m p o r t a n t explanatory variable. While the coefficient o n this a d d i t i o n a l variable is significantly dif ferent f r o m zero i n the regressions covering the f u l l p e r i o d , i t is n o t signi f i c a n t l y different f r o m zero i n 3 o f the 4 regressions for the post-EMS sub-p e r i o d .
Thus, the evidence o n long-run homogeneity is somewhat m i x e d . I t should be remembered that the price series used are for t o t a l m a n u f a c t u r i n g i n d u s t r y . Differences i n the weights attached t o different components w i l l lead t o divergences f r o m relative PPP i f the price trends for these different com ponents diverge.
4.4 Tests of Short-term PPP
We have already seen one set o f F-tests w h i c h indicates that the response to exchange rate changes is slower than that t o prices. Further tests c o n f i r m this as a strong result. F o r m u l a t i o n s w h i c h impose an identical response i n each quarter b y Irish prices t o changes i n foreign currency prices and exchange rates are rejected against those w h i c h a l l o w differential effects. This result holds very generally across countries, sectors and sub-periods. For example, the F-test for imposing an i d e n t i c a l response r e s t r i c t i o n o n E q u a t i o n (5) yields a value o f 3.65 as against a 1 per cent critical value F ( 8 , 2 ) o f 3 . 4 1 . These tests c o n f i r m the result o f L e d d i n (1988) t h a t purchasing p o w e r p a r i t y , expressed o n a p e r i o d b y p e r i o d basis, can be rejected: they also indicate, however, t h a t Leddin's i m p o s i t i o n o f i d e n t i c a l coefficients o n the foreign price and exchange rate terms is rejected b y the data.
this r e s t r i c t i o n on E q u a t i o n (5) yields a value o f 1.28 as against a critical value F ( 2 , 2 3 ) o f 3.42.
The response t o German price changes i n D M terms tends t o be larger (and better defined) than the response to exchange rate changes. I n the case o f the U K the opposite is the case w i t h the exchange rate terms being generally better defined than the price terms. The short-term response t o exchange rate changes tends t o be somewhat slower than t o foreign currency prices.
4.5 Tests of Asymmetric Response to Changes in Foreign Prices
Baker et al. (1986) suggested that domestic prices w o u l d respond faster t o a rise i n the price o f foreign currency than t o a fall. The possibility o f a stronger response t o exchange rate induced increases i n c o m p e t i t o r s ' prices (in Irish p o u n d terms) than t o decreases was tested as follows. First, a d u m m y variable was constructed t a k i n g the value 1 for an increase i n the U K exchange rate and zero otherwise. A n a d d i t i o n a l variable was f o r m e d by m u l t i p l y i n g the short-term exchange rate changes by this d u m m y variable, thus giving a variable equal t o the exchange rate change i f this was positive, b u t zero o t h e r w i s e .1 7
A positive significant coefficient o n this variable w o u l d indicate a more r a p i d response t o positive exchange rate changes. The results show that the coef ficients o n the d u m m y variables, while all bearing the expected positive sign, i m p l y i n g a faster response t o a depreciation o f the Irish p o u n d , are j o i n t l y n o t significant f r o m zero. However, due t o the small number o f observations for an appreciation o f the Irish p o u n d against Sterling, this test is n o t conclusive.
4.6 Non-nested Tests against Trade-weighted Foreign Price Variables Trade weights for the U K , Germany and the US move fairly steadily f r o m 0.75, 0.13, 0.12, respectively, t o 0.55, 0.20, 0.25 over the p e r i o d 1975 t o 1987. We have already seen that the i m p l i e d weight on German prices i n the aggregate level regressions is m u c h greater than even this end-period weight, and the i m p l i e d weight on the U K is correspondingly lower. However, the fact that the trade weights move over the p e r i o d to reflect Germany's increas i n g trade share c o u l d give a trade-weighted approach some advantages over the f i x e d weights o f the regression approach. Simple trade weights were used t o create w e i g h t e d averages o f U K , German and US prices and exchange rates: the results r e p o r t e d b e l o w do n o t suggest that more sophisticated weighting sys tems w o u l d lead t o different conclusions. Non-nested tests provide the appro priate f r a m e w o r k for choosing between the trade-weighted and regression-w e i g h t e d approaches.
The results indicate t h a t the regression-weighted approach encompasses the results o f the trade-weighted approach. The trade-weighted regressions are rejected against a j o i n t m o d e l i n c o r p o r a t i n g separate i n f o r m a t i o n o n the U K and Germany (the test statistic o f 4.72 exceeding the critical F ( 1 8 , 1 5 ) at the 1 per cent level o f 3.43), w h i l e the regression-weighted approach is n o t rejected against a j o i n t m o d e l i n c l u d i n g trade-weighted i n f o r m a t i o n (the test statistic o f 1.31 being b e l o w the 5 per cent c r i t i c a l F ( 9 , 1 5 ) o f 2.59). The r e p o r t e d statistics are based o n the inclusion o f three lags o f each price and exchange rate variable, as i n E q u a t i o n ( 5 ) , b u t similar results were o b t a i n e d w i t h m o r e restricted specifications.
Specifications w h i c h use U K data alone t o p r o x y foreign prices are also strongly rejected against specifications i n c l u d i n g German price data. Further m o r e , the specifications i n v o l v i n g the U K alone y i e l d unstable and counter i n t u i t i v e results, i n c l u d i n g rejection o f long-run homogeneity, and a sign o n the error c o r r e c t i o n t e r m w h i c h w o u l d lead t o Irish prices diverging f r o m those i n the U K .
There w o u l d seem t o be t w o m a i n reasons w h y the influence o f German prices exceeds their trade weight. The first is t h a t German prices m a y t o some extent p r o x y prices for France, the Netherlands and other European countries n o t i n c l u d e d i n the analysis. I t is d o u b t f u l , however, i f this is the w h o l e story. A n alternative, m e n t i o n e d earlier, is that German prices exert an influence o u t o f p r o p o r t i o n t o their trade w e i g h t , possibly because German firms are regarded as price leaders; further investigation w o u l d be needed t o establish whether or n o t this is the case.
V R E V I E W O F D I S A G G R E G A T E D R E S U L T S
We also applied the m o d e l described above i n E q u a t i o n (4) t o the price data for the o u t p u t o f a range o f n o n - f o o d sectors o f m a n u f a c t u r i n g i n d u s t r y .1 8
I n each case data for prices and exchange rates i n the U K , Germany and the US, were t r i e d together w i t h wage rates. I n the case o f certain sectors price and exchange rate data for some other possibly relevant countries were t r i e d . The results o f these regressions are summarised i n Table 4 .1 9 The same dynamic
specification was used i n each case w i t h no lags on the price change variables and 3 lags o n the exchange rate changes. I n all b u t one case a specification i n c l u d i n g 3 lags o n the change i n prices was rejected at the 5 per cent level i n favour o f this m o r e restricted specification used here.
18. See Callan and FitzGerald (1988) for full details of these results.
While the R2 statistics r e p o r t e d for these disaggregated equations are t y p i
cally l o w e r t h a n t h a t for the aggregate equation, they are all highly significant, and compare favourably w i t h other t i m e series analyses o f dependent variables differenced i n logarithms. A t such a disaggregated level the increasing i m p o r tance o f special factors, n o t captured b y the regressors, m a y account for the somewhat l o w e r R2 statistics r e p o r t e d .
The 11 equations for the different sectors o f m a n u f a c t u r i n g cover the b u l k of the n o n - f o o d o u t p u t o f t h a t sector (over 60 per cent o f net o u t p u t ) . One major omission is the electronics sector w h i c h is also o m i t t e d f r o m the index covering t o t a l m a n u f a c t u r i n g . As can be seen f r o m the table the results are reasonably consistent w i t h the results obtained f r o m the aggregate equation.
I n eight sectors at least one o f the 6 coefficients are significant i n d i c a t i n g a significant long-term relationship between Irish prices and the explanatory variables. Domestic costs, p r o x i e d b y wage costs, are o f o n l y l i m i t e d i m p o r
tance i n d e t e r m i n i n g prices; they are o n l y significant i n t w o sectors i n the l o n g r u n .2 0 Thus, for m o s t sectors the SOE m o d e l is appropriate. As w i t h the
results presented i n Section 4 , German prices generally have a higher weight i n determining Irish prices than do U K prices. The results c o n f i r m the slower speed o f adjustment o f I r i s h prices t o changes i n exchange rates than t o changes i n foreign currency prices. F i n a l l y , these results c o n f o r m e d t o the results i n Section 4 i n ascribing l i t t l e or n o weight t o other countries' p r i c e s2 1
i n explaining Irish prices.
T a b l e 4 : Results for Individual Industrial Sectors
Sector 1 2 3 4 5 6 72 2
5 9 10 11
a 0.788 1.510 1.139 1.101 1.171 2.003 0.567 0.404 - 0 . 0 5 6 0.665 0.493 (3.6) (3.6) (1.4) (4.8) (4.0) (3.4) (1.9) (2.3) (0.3) (3.4) (1.4)
^OGR 0.261 0.117 0.014 1.005 - 0 . 0 0 7 0.176 0.270 0.744 (0.7) (0.1) (0.0) (3.9) (0.0) (0.9) (1.8) (5.4)
70 G R 0.223 -0.010 0.074
0.107 0.371 - 0 . 0 0 4 - 0 . 0 3 8 0.031
70 G R
(3.4) (0.1) (1.5) (2.1) (2.2) (0.2) (0.3) (0.3) 0.069 0.140 0.171 0.233 0.511 - 0 . 0 2 0 0.281 0.415 (1.0) (1.9) (3.2) (4.1) (2.9) (0.0) (2.1) (3.9)
^2GR 0.002 0.037 0.160 0.081 0.173 0.077 - 0 . 3 0 8 0.216 (0.0) (0.4) (2.9) (1.5) (1.0) (2.1) (2.4) (2.0)
^ G R 0.134 - 0 . 1 0 7 0.015 0.162 0.200 0.032 0.188 0.081 ^ G R
(2.3) (1.3) (0.3) (3.1) (1.2) (1.0) (1.4) (0.8)
0.436 0.616 0.339 0.293 - 0 . 0 8 8 0.183 -0.010 0.853 0.463 (2.8) (2.7) (2.1) (1.4) (0.4) (0.3) (0.0) (3.9) (3.8)
7O U K 0.018 0.081 0.076 - 0 . 0 6 3 0.109 0.106 0.090 0.155 0.019 7O U K
(0.2) (1.7) (1.6) (1.6) (2.8) (0.8) (2.4) (3.1) (0.4)
^ l U K 0.232 0.146 0.099 0.054 - 0 . 0 4 3 0.026 0.026 0.190 0.105 ^ l U K
(3.1) (3.2) (2.5) (1.3) (1.0) (0.2) (0.7) (3.8) (2.1)
72 U K 0.086
0.079 0.012 0.023 0.033 0.124 0.075 - 0 . 0 2 0 0.096
72 U K
(1.0) (1.7) (0.2) (0.5) (0.8) (0.9) (2.0) (0.3) (1.8)
T3UK. 0.072 0.111 - 0 . 0 2 2 0.168 0.040 0.026 0.062 0.026 0.091 (0.9) (2.4) (0.5) (4.3) (1.0) (0.2) (1.6) (0.5) (1.9)
eo
0.037 0.015 - 0 . 0 8 6 0.072
(1.3) (0.7) (2.1) (2.7)
0.012 - 0 . 0 0 2 - 0 . 0 6 8 0.112
(0.4) (0.1) (1.3) (3.6)
e2
0.024 0.041 0.124 0.087
(0.7) (1.5) (2.0) (2.4)
e3 0.076 0.040 0.021 - 0 . 0 6 7
(2.2) (1.5) (0.4) (1.4)
^GR 0.131 0.126 0.114 0.099 0.126 0.346 0.055 0.097 0.076 (2.6) (3.1) (1.5) (3.7) (4.2) (3.1) (1.8) (3.3) (1.4)
0.133 0.050 0.084 0.068 - 0 . 0 4 9 0.000 0.086 - 0 . 0 6 7 (3.3) (0.7) (3.8) (2.4) (1.0) (0.0) (1.6) (1.4)
ew 0.042 0.051 - 0 . 0 1 4 0.079
(1.4) (2.4) (0.3) (2.5)
R2 0.647 0.664 0.706 0.685 0.756 0.393 0.453 0.694 0.785 0.292 0.472
Sectors: 1. Chemicals 2. Metal Products 3. Electrical Engineering 4. M o t o r Vehicles and Parts
5. Textiles
6. Leather and Footwear 7. Clothing
8. Timber and Furniture
9. Paper and Paper Products 10. Rubber and Rubber Products 11. Processing o f Plastics
V I C O N C L U S I O N S
The results presented i n this paper show t h a t the SOE m o d e l o f price d e t e r m i n a t i o n is appropriate t o Irish manufacturing o u t p u t . Domestic wage costs play o n l y a m i n o r role i n price d e t e r m i n a t i o n , while foreign price i n f l u ences are d o m i n a n t . I n this respect, the results c o n f i r m the consensus o f earlier research o n this t o p i c ; b u t the results on the nature and t i m i n g o f the foreign influence are a good deal more novel.
Short-run h o m o g e n e i t y is rejected b y the data. While long-run homogeneity is generally rejected for the f u l l data sample, i t is accepted for the post-EMS p e r i o d .
Generally the effects o f changes i n exchange rates on prices are m u c h slower to materialise than are the effects o f changes i n foreign currency prices. This difference i n behaviour can be explained b y the fact that firms are reluctant to change their o u t p u t prices very frequently. As a result, they have t o set future prices based o n their expectations o f the Irish p o u n d price o f foreign c o m p e t i t o r s . When foreign c o m p e t i t o r s raise their prices i n foreign currency terms there is l i t t l e l i k e l i h o o d t h a t they w i l l subsequently reduce t h e m , so domestic producers w i l l be quick t o f o l l o w suit. However, i f foreign prices i n Irish p o u n d terms change because o f exchange rate changes, past experience suggests, especially i n the case o f the Sterling-Irish p o u n d exchange rate, that there is quite a p o s s i b i l i t y that the change may be reversed i n the immediate future. As a result, firms are l i k e l y t o be slower t o change their expectations about future values o f the exchange rate as a result o f a single quarter's f i g u r e s .2 3
There is no evidence o f an asymmetric response to appreciation and depreci a t i o n vis-a-vis Sterling. However, the number o f observations available for the post-EMS p e r i o d are s t i l l t o o few to p r o p e r l y test this hypothesis. I n any event, this hypothesis may be more appropriately applied t o a m o d e l o f the f o r m a t i o n o f domestic consumer prices, as suggested by Baker et al., 1986.
Germany has a stronger influence i n determining Irish prices than the trade weights suggest. I n the post-EMS p e r i o d its role i n o u t p u t price d e t e r m i n a t i o n is r o u g h l y equal t o that o f the U K . The p a t t e r n o f response o f Irish prices t o foreign prices was significantly different f r o m that suggested b y an approach based o n trade weights.
Because o f a l i m i t e d number o f consistent observations for the pre-EMS p e r i o d i t is d i f f i c u l t t o test for a change i n p a t t e r n o f behaviour f o l l o w i n g the break i n the l i n k w i t h sterling consequent on EMS membership. However, the evidence f r o m the aggregate price series does indicate a change i n p r i c i n g
behaviour i n 1979 w i t h an increase i n the i m p o r t a n c e o f German prices and a fall i n the i m p o r t a n c e o f U K p r i c e s .2 4 I t is d i f f i c u l t t o i d e n t i f y a change i n
short-run p r i c i n g behaviour. The differential response t o exchange rate changes and foreign currency price changes has also taken o n added importance n o w that the Sterling-Irish p o u n d rate is flexible.
These results have a n u m b e r o f i m p l i c a t i o n s for p u b l i c p o l i c y . I n so far as the p r i c i n g behaviour o f Irish firms reflects their views as t o w h o their com petitors are, these results suggest a change i n the t r a d i t i o n a l trade-weighted approach t o measuring competitiveness. O n the basis o f these results, EMS countries, p r o x i e d b y Germany, should have a m u c h bigger weight i n c o m petitiveness measures t h a n w o u l d be suggested b y t r a d i t i o n a l trade weights.
The slower response o f domestic prices t o changes i n the exchange rate than t o changes i n foreign currency prices has i m p l i c a t i o n s for the Irish eco n o m y o n the c o m p l e t i o n o f the EC m a r k e t i n 1992. I f the U K remains o u t of the EMS u n c e r t a i n t y about the bilateral Sterling-Irish p o u n d exchange rate w i l l pose problems i n price d e t e r m i n a t i o n . W i t h o u t any customs or fiscal barriers the slow adjustment o f prices t o exchange rate changes c o u l d give rise to large t e m p o r a r y trade flows. The alternative is that either prices i n I r e l a n d change m o r e r a p i d l y i n the face o f changes i n the Sterling-Irish p o u n d exchange rate or else the v a r i a b i l i t y o f U K prices i n Sterling terms increases.
REFERENCES
B A K E R , T . , 1 9 8 5 . " T r e n d s i n M a n u f a c t u r i n g O u t p u t a n d Wage C o s t s 1 9 8 0 - 1 9 8 4 " , A p p e n
d i x to Quarterly Economic Commentary, A p r i l , p p . 2 6 - 3 5 .
B A K E R , T . , S . S C O T T , a n d T . Q U I N N , 1 9 8 6 . Quarterly Economic Commentary, A p r i l .
B R A D L E Y , J . , 1 9 7 7 . " L a g s i n the T r a n s m i s s i o n of I n f l a t i o n " , The Economic and Social
Review, Vol. 8 , p p . 1 4 9 - 1 5 4 .
C E N T R A L S T A T I S T I C S O F F I C E , 1 9 8 8 . Census of Industrial Production, 1 9 8 5 , D u b l i n :
cso.
C A L L A N , T . , a n d J . F I T Z G E R A L D , 1 9 8 8 . "Price D e t e r m i n a t i o n i n I r e l a n d : E f f e c t s of changes i n exchange rates a n d exchange rate regimes", D u b l i n : E S R I M e m o r a n d u m
Series N o . 1 7 9 .
D I C K E Y , D . A . , a n d W . A . F U L L E R , 1 9 7 9 . " D i s t r i b u t i o n of the E s t i m a t e s for A u t o -regressive T i m e Series w i t h a U n i t R o o t " , Journal of the American Statistical Associ
ation, V o l . 7 9 , p p . 4 2 7 - 4 3 1 .
D I C K E Y , D . A . , a n d W . A . F U L L E R , 1 9 8 1 . " T h e L i k e l i h o o d R a t i o Statistics for A u t o
-regressive T i m e Series w i t h a U n i t R o o t " , Econometrica, V o l . 4 9 , p p . 1 0 5 7 - 1 0 7 2 .
E N G L E , R . F . , a n d C . W . G R A N G E R , 1 9 8 7 . " C o - I n t e g r a t i o n a n d E r r o r C o r r e c t i o n :
R e p r e s e n t a t i o n , E s t i m a t i o n a n d T e s t i n g " , Econometrica, V o l . 5 5 , N o . 2, p p . 2 5 1 - 2 7 6 . E N G L E , R . F . , a n d B . S . Y O O , 1 9 8 7 . " F o r e c a s t i n g a n d T e s t i n g in C o - I n t e g r a t e d S y s t e m s " ,
Journal of Econometrics, V o l . 3 5 , N o . 1, p p . 1 4 3 - 1 5 9 .
F I S H E R , F . M . , 1 9 7 0 . "Tests o f E q u a l i t y b e t w e e n Sets of Coefficients i n T w o L i n e a r
Regressions: A n E x p o s i t o r y N o t e " , Econometrica, V o l . 3 8 , p p . 3 6 1 - 3 6 6 .
F I T Z G E R A L D , J . , 1 9 8 3 . " A S t u d y of the D e t e r m i n a n t s of I r i s h I n d u s t r i a l O u t p u t P r i c e s " , E S R I seminar paper.
F R E N K E L , J . A . , 1 9 8 1 . " T h e C o l l a p s e of P u r c h a s i n g P o w e r Parities during the 1 9 7 0 s " ,
European Economic Review, V o l . 16, p p . 1 4 5 - 1 6 5 .
G E A R Y , P . T . , 1 9 7 6 a . " L a g s i n the T r a n s m i s s i o n of I n f l a t i o n : S o m e P r e l i m i n a r y E s t i
m a t e s " , The Economic and Social Review, V o l . 7, p p . 3 8 3 - 3 8 9 .
G E A R Y , P . T . , 1 9 7 6 b . " W o r l d Prices a n d the I n f l a t i o n a r y Process in a S m a l l O p e n E c o n o m y — T h e C a s e of I r e l a n d " , The Economic and Social Review, V o l . 7, p p . 3 9 1 - 4 0 0 .
G E A R Y , P . T . , a n d C . M c C A R T H Y , 1 9 7 6 . "Wage a n d P r i c e D e t e r m i n a t i o n i n a L a b o u r E x p o r t i n g E c o n o m y : T h e Case o f I r e l a n d " , European Economic Review, V o l . 8, N o . 3,
p p . 2 1 9 - 2 3 4 .
G O L D S T E I N , M . , 1 9 7 4 . " T h e E f f e c t s of E x c h a n g e R a t e Changes o n Prices a n d Wages i n
the U K : an E m p i r i c a l S t u d y " , IMF Staff Papers, V o l . 2 3 , N o . 4.
G R A N G E R , C . W . , 1 9 8 6 . " D e v e l o p m e n t s i n the S t u d y of C o i n t e g r a t e d E c o n o m i c V a r i ables", Oxford Bulletin of Economics and Statistics, V o l . 4 8 , N o . 3, p p . 2 1 3 - 2 2 8 .
H E N D R Y , D . F . , 1 9 8 3 . " E c o n o m e t r i c Modelling — the C o n s u m p t i o n F u n c t i o n i n R e t r o
s p e c t " , Scottish Journal of Political Economy, V o l . 3 0 , N o . 3, p p . 1 9 3 - 2 0 0 .
H O N O H A N , P . , a n d J . F L Y N N , 1 9 8 6 . " I r i s h I n f l a t i o n i n E M S " , The Economic and
Social Review, V o l . 17, p p . 1 7 5 - 1 9 1 .
I S A R D , P . , 1 9 7 7 . " H o w F a r C a n We P u s h the L a w of O n e P r i c e ? " , American Economic
Review, V o l . 6 7 , p p . 9 4 2 - 9 4 8 .
K R A S K E R , W . S . , E . K U H , a n d R . E . W E L S C H , 1 9 8 3 . " E s t i m a t i o n for D i r t y D a t a a n d F l a w e d M o d e l s " , i n Z . G r i l i c h e s a n d M . D . Intriligator (eds.), Handbook of Econo
L E D D I N , A . , 1 9 8 8 . "Interest a n d Price P a r i t y a n d F o r e i g n E x c h a n g e M a r k e t E f f i c i e n c y : T h e I r i s h E x p e r i e n c e i n the E u r o p e a n M o n e t a r y S y s t e m " , The Economic and Social
Review, V o l . 19, p p . 2 1 5 - 2 3 1 .
R E S T R I C T I V E P R A C T I C E S C O M M I S S I O N , 1 9 8 8 . Report of the Restrictive Practices
Commission, D u b l i n : S t a t i o n e r y O f f i c e .
T H O M , R . , 1 9 8 9 . " R e a l E x c h a n g e R a t e s , Co-integration a n d P u r c h a s i n g P o w e r P a r i t y :
A p p e n d i x 1 : Calculated Values of the Test Statistics Underlying Figure 1.
Specifications 1, 2 and 3 and dynamics A , B , C and Z are as described i n Figure 1 and the related t e x t .
Z l F ( l l , 1 7 ) = 1.05 Z 2 F ( 2 , 2 8 ) = 8 . 0 4 * * Z 3
F ( 6 , l l ) = 7 . 2 9 * * F ( 6 , 2 2 ) = 8 . 0 5 * * F ( 6 , 2 4 ) = 8 . 1 2 * *
A l F ( 1 1 , H ) = 1.61 A 2 F ( 2 , 2 2 ) = 7 . 2 7 * * A 3
F ( 6 , l l ) = 2 . 2 6 F ( 6 , 2 2 ) = 1 . 8 7 F ( 6 . 2 4 ) = 0 . 5 0
B l F ( l l , 1 7 ) = 1.18 B 2 F ( 2 , 2 8 ) = 3 . 3 1 + B 3
F ( l , 1 7 ) = 0 . 5 0 F ( l , 2 8 ) = 1 . 6 5 F ( l , 3 0 ) = 5 . 6 0 *
C I F ( l l , 1 8 ) = =1.33 C 2 F ( 2 , 2 9 ) = 5 . 5 9 * * C 3
A test statistic w h i c h is significant at 1 per c e n t ( * * ) , 5 per cent(*) and 10 per cent(+) indicates that the less general specification (i.e., the one w i t h more restrictions and, hence, fewer estimated parameters) is rejected against the more general specification. For example, since Z is less general than A , the test statistic F ( 6 , l l ) = 7 . 2 9 indicates that Z l is rejected against A l . Specifica t i o n A is m o r e general than B w h i c h is m o r e general than C while dynamics 1 are m o r e general than 2 w h i c h are more general than 3.
T y p e s e t by W e n d y A . C o m m i n s , T h e C u r r a g h ; M a k e - u p by Paul B r a y S t u d i o .