Real Exchange Rate Fluctuatons, Wage Stckness and Markup Adjustments Yothn Jnjarak and Kanda Nakno Nanyang Technologcal Unversty and Purdue Unversty January 2009 Abstract Motvated by emprcal evdence on the role of wage stckness n nflaton dynamcs, ths paper studes the relatonshp between tradablty and dstrbuton of ndustry-specfc real exchange rate fluctuatons. We frst offer a smple model where labor s the only nput for the nontraded ntermedate good, and the fnal good n each ndustry s produced from the nontraded nput and an mported ntermedate good. Usng the quarterly data from the US and Canada, we decompose movements of ndustry-specfc real exchange rates nto relatve wage adjustments and markup adjustments. The decomposton hghlghts the mportance of sectoral heterogenety of wage stckness and also tradablty of goods as the determnants of sector-specfc real exchange rates. JEL codes: F41, F42 Keywords: real exchange rate; wage stckness; markup; Dvson of Economcs, Nanyang Technologcal Unversty, S3-B2A-06 Nanyang Ave., Sngapore 639798. Emal: yjnjarak@ntu.edu.sg. Department of Economcs, Purdue Unversty, West Lafayette, IN 47907, USA. Emal: knakno@purdue.edu. 1
1 Introducton Recent lterature on the real exchange rate RER) has found that tradablty of goods plays an mportant role on both tme-seres and cross-secton dstrbuton of RER. Specfcally, Crucn, Telmer and Zacharads 2005) found that the cross-sectonal dstrbuton of sector-specfc RER s attrbuted to tradablty of ntermedate nputs rather than tradablty of fnal good. Betts and Kehoe 2006) found that RER volatlty s decreasng n tradablty of goods. Fnally, Nakno 2008) demonstrated that endogenous tradablty of goods can explan the varance decomposton of RERs. Ths study attempts to nvestgate the effects of tradablty of goods on the dstrbuton of sector-specfc RERs usng the U.S. and Canadan data. The unque feature n our study s that we vew nontradablty of labor as the major determnant of tradablty of nput. We ask two followng questons. Frst, how hgh s the degree of sectoral heterogenety of movements of sector-specfc RER? Second, s the sectoral heterogenety drven by the heterogenety of wage stckness or the heterogenety of retaler s markup adjustments? Conceptually, fnal goods are produced from traded and nontraded ntermedates such as offce rental servce, marketng and dstrbutng servces. For ths reason, our model assumes that labor s the sole nput for the nontraded ntermedate good. The fnal good n each ndustry s assumed to be produced from the nontraded nput and an mported ntermedate good. To allow for the role of markup prcng documented wdely n the emprcal lterature, we assume that the mporters have monopoly power when they sell the mported ntermedate goods to the fnal good producers. Lkewse, there s a monopoly retaler who dstrbutes the fnal good n each ndustry to consumers. Gven ths structure, we can decompose movements of sector-specfc RER nto two components as follows: relatve wage adjustments n the two countres; and relatve markup adjustments by sellers of ntermedate goods and fnal goods. In our dataset we observe sector-specfc prces and wages, hence we can calculate the markup adjustments as the resduals. Then we calculate the contrbuton of each components nto the 2
sector-specfc RERs. The methodology s smlar to that n the semnal work on RER varance decomposton by Engel 1999). Our work has drect mplcatons to theoretcal work on RER. It s the frst study that provdes stylzed facts about the mportance of heterogenety of wage stckness comparng to the mportance of heterogenety of markup adjustments n RER fluctuatons. The exstng emprcal work has documented mportance of wage stckness as a determnant of nflaton dynamcs, but none has nvestgated ts mportance as a determnant of RER dynamcs Taylor 1999; Huang and Lu 2002). Secton 2 outlnes the model and provdes the framework for the varance decomposton of sector-specfc RERs. The descrpton of data are n Secton 3. 2 The Model For each ndustry, the retaler dstrbutes the fnal good wth markup prcng. subscrpt t denote the perod. Let P t = µ tm t, 1) where P t s the retal prce, µ t s the retaler markup, and M t s the producer prce. The retaler buys the fnal good from a producer, who produces t from an ntermedate nput and labor wth Cobb-Douglas technology. M t = φ t W t ) α Et τ C t) 1 α, 2) where φ t s the producer markup, W t s wage, α s the share of nontraded nput, E t s the unts of home currency per unt of source country s currency, τ, τ > 1)s gross tradecosts, and C t s the unt cost of mport n source country s currency. Suppose that the same producton structure holds true for foregn frms, and they also mport ntermedate 3
goods from the same source country. Defne the sector-specfc RER or Q t as follows. Q t = S tpt, Pt where the superscrpt denotes the foregn varables and S t s the unts of home currency per unt of foregn currency. Hence, S t = E t /E t. Assume that technology s symmetrc n the home and foregn economes. The equlbrum sector-specfc RER becomes: µ t φ t Wt ) α Et τ Ct) 1 α Q t = E t Et µ tφ t Wt ) α E t τ Ct) 1 α ) ) µ = t φ t St W α ) t τ 1 α 3) µ tφ t Wt τ Defne the markup term as H t = µ t φ t )/µ tφ t). dfferencng 3) gves the movement of sector-specfc RER as follows. Takng natural logarthm and frst q t = h t + α st + w t w t), 4) where lowercase denotes natural logarthm and denotes the frst dfference. If we observe exchange rate seres, prce seres and wage seres, we can calculate the markup term as resduals. Fnally, we calculate the contrbuton of relatve wage nto the overall RER varance as follows. vw = var s t + wt where var denotes varance and cov denotes covarance. wt) + cov s t + wt wt, h t), 5) var qt) 4
3 The Emprcs The data are quarterly from 1982 to 2005. The ndustres are classfed by the NAICS code. The prces seres are producer prce ndex PPI) compled by Bureau of Labor Statstcs BLS) and the Canada s Natonal Statstcal Agency. The wage seres are the employment cost ndex compled by the BLS and the Canada s Natonal Statstcal Agency. The exchange rate data are from the Internatonal Fnancal Statstcs. The sector-specfc share of nontraded nput comes from the U.S. nput-output table. References Betts, C. M. and T. J. Kehoe 2006). U.S. real exchange rate fluctuatons and relatve prce fluctuatons. Journal of Monetary Economcs 53 7), 1297 1326. Crucn, M. J., C. I. Telmer, and M. Zacharads 2005). Understandng European real exchange rates. Amercan Economc Revew 95 3), 724 738. Engel, C. 1999). Accountng for U.S. real exchange rate changes. Journal of Poltcal Economy 107 3), 507 538. Huang, K. X. and Z. Lu 2002). Staggered prce-settng, staggered wage-settng and busness cycle persstence. Journal of Monetary Economcs 49, 405 433. Nakno, K. 2008). Real exchange rate fluctuatons, endogenous tradablty and exchange rate regmes. Journal of Monetary Economcs 55 3), 645 663. Taylor, J. B. 1999). Staggered prce and wage settng n macroeconomcs. In J. B. Taylor and M. Woodford Eds.), Handbook of Macroeconomcs, pp. 1009 1050. Elsever. 5