關稅訊息的總體效果 - 政大學術集成

全文

(1)國立政治大學社會科學院經濟學系碩士學位論文 Department of Economics College of Social Sciences National Chengchi University. Macroeconomic Effects of News on Tariff. 政治大立關稅訊息的總體效果. ‧. ‧ 國. 學. Nat. n. sit. Liu, Chih-Cheng. Ch. engchi. er. io. al. y. 劉至誠. i n U. v. 指導教授 : 黃俞寧博士 Advisor : Hwang, Yu-Ning, Ph.D. 中華民國 104 年 7 月 July, 2015.

(2) 摘要. 本文建構一個小型開放的經濟體系，價格具有僵固性的 DSGE 模型來探討提早釋出的關稅調降訊息所造成總體經濟面的影響。在政府簽訂貿易協定來降低關稅的訊息藉由媒體釋出時，同時民眾具理性預期的假設下，家計單位會因這樣的訊息而改變對未來的預期，在關稅還未實際調降時便改變行為決策。我們研究在不同的協商可能結果: (1)談判破裂，貿易協定未如預期簽訂。(2)談判成功，但貿易協定簽訂的結果比當初訊息所透露的降得更多或較少。(3)談判成功，並完全實現當初訊息的內容。(4)簽訂的時間高於預期，導致實行的時間延期。從結果中我們發現，提早釋出的關稅訊息會抑制. 政治大絡，會等到真正調降的時後才出現。所以越早釋出關稅調降的訊息，會造成經濟體系所立. 民眾的消費和投資，在短期會立即造成需求面的負向衝擊。而降低關稅所帶來的市場活. 需付出“等待政策實行的成本”越大。. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. i n U. v. 關鍵字：關稅；訊息；動態隨機一般化均衡；小型開放經濟；價格僵固.

(3) Abstract There is global trend of economic integration across the world by removing the trade barrier. While the free trade agreements normally include the tariff reduction, the negotiations of the agreement may take a long time, and in some cases, the negotiation may fail. Therefore, the tariff reduction’s effects on the economy can be different if it is realized as expectation or not. With a small open economy DSGE model, this paper examines the effects of news preannouncement on tariffs. With the assumption of rational expectation, households will change their expectation when the news of tariff decrement is preannounced. However, whether or not the news on tariffs can be realized as expectation will lead to different dynamics. In this. 政治大 reduction is not realized (2) If negotiation 立 succeeds, but the amount of actual decrement is more. study, we consider various plausible scenarios: (1) If negotiation fails, thus the news on tariff. ‧ 國. 學. or less than people originally expected. (3) Negotiation succeeds and the content of news is fully realized. (4) Negotiation succeeds, but it takes more time than expected to be realized, thus the. ‧. policy implementation is postponed. Our study reveals that the news preannouncement restrains. sit. y. Nat. consumption and investment before the negotiation is completed, and leads to a negative impact. n. al. er. io. on the economy in the short run. The benefits of the tariff decrement appear if the policy is. i n U. v. actually implemented as expected. However, the economy suffers more if the negotiation takes. Ch. too much time after the news is released.. engchi. Key words: Tariff; News; DSGE; Small open economy; Nominal rigidity.

(4) Content. 1. Introduction ............................................................................................................................................................2 2. Literature review ....................................................................................................................................................3 3. Model ......................................................................................................................................................................7 3.1 Goods market ....................................................................................................................................................7 3.2 Firms .................................................................................................................................................................7 3.2.1 Final goods producer .................................................................................................................................7 3.2.2 Intermediate goods firms ...........................................................................................................................9 3.3 The representative household .........................................................................................................................12. 政治大 3.5 Central bank....................................................................................................................................................16 立 3.6 Market cleaning condition ..............................................................................................................................16. 3.4 Tariff and government ....................................................................................................................................15. ‧ 國. 學. 3.7 Exogenous variables .......................................................................................................................................17 4. Calibrated parameters ...........................................................................................................................................17. ‧. 5. Analysis ................................................................................................................................................................19. y. Nat. sit. 5.1 Case 1: Tariff decrement with and without preannouncement .......................................................................20. er. io. 5.2 Case 2: The time news on tariff reduction is released before implementation ...............................................21. al. n. v i n Ch U 5.4 Case 4: The tariff reduction is realized unexpectedly or is realized at the size higher or e n g(news c h iis unrealized. 5.3 Case 3: Policy delay .......................................................................................................................................21. lower than expectation) ........................................................................................................................................22 6. Conclusion ............................................................................................................................................................22 Reference ..................................................................................................................................................................24 Table .........................................................................................................................................................................26 Figure ........................................................................................................................................................................28 Appendix ..................................................................................................................................................................35. 1.

(5) 1. Introduction. The rapid pace for the integration of the world economy has led to more and more trade agreements. With the global trend of integration, many countries strive to win the trade partners by signing free trade agreements to enhance their competitiveness in international trades. The free trade agreement (FTA) coverage rate has become a reference index to evaluate the national trade competitiveness. However, there are tradeoffs. The advantage of free trade agreements is to increase the trade amounts by allowing countries to specialize in the industries with comparative advantage. However, this may lead to the losses of the industries lacking. 政治大 normally take long for the government to raise the benefits and reduce the losses. It is 立 comparative advantage and may result in protests. Therefore, the negotiations between countries. particularly a main concern for a small open economy which may have weaker bargaining power.. ‧ 國. 學. There have been many research puts efforts on the impact of the FTA on the economy and. ‧. welfare after the agreement is signed (Bagwell and Staiger 1999, Eaton and Kortum 2002 and. sit. y. Nat. Caliendo and Parro 2015). However, since the negotiation process may take a long time, the. io. er. agreements can be signed long after the government makes the announcement of its attempt to start the negation. The expectation of the likely lower tariffs may affect the decisions of. al. n. v i n Ctheh free trade agreement consumers and profits of firms before is signed. . The time lag between engchi U 1. the announcement and the signing of the FTA may result in different economic performance. In particular, whether or not the agreement can be signed, or if the reduction of tariff is same as expectation may cause different results. In spite of the abundant literature on tariffs, there is no evaluation of the news on tariff on the macroeconomic dynamics. Therefore, based on recent works for news shock (or expectation shock) where the expectation for the future, given the news announced, can drive the business cycles, we want to. The expectation of the FTA’s impacts may lead to economic and social responses. For example, in 2009, twenty thousand framers greeted Barack Obama with anger for the FTA between Korea and United states before it is signed in 2011 and take effect in 2012. It is because farmers do not think there is enough protection of the domestic agriculture. 2 1.

(6) examine the effects of the news on tariffs on the macro fluctuations, and how the dynamics may change if the negotiation turns out to fail. With a small open economy dynamic stochastic general equilibrium (DSGE) model with the news on tariff, we study various scenarios. We discuss the case where the agreement is out of expectation which include the degree of adjustment in outcome is higher or lower than the news announced before and what if negotiation turns into failure then leads to no agreement. For the case which the news is realized as expected, we compare the effects such as implement postponed and different length of time for preannouncement. This study may provide some policy implications for the government.. 2. Literature review. 立. 政治大. This paper examines the macroeconomic effects of tariff news in an intertemporal small. ‧ 國. 學. open economy model with price rigidities. The first macroeconomic study on tariffs can be. ‧. traced back to Mundell (1961), who refers to the imposition of tariffs as a kind of commercial. sit. y. Nat. policy that is less effective than both monetary and fiscal policies. Therefore, the imposition of. io. flexible exchange rate regime.. er. tariff will cause the deflation and contractionary output and higher unemployment under the. al. n. v i n C hin Mundell (1961),UEichengreen (1981) tries to examine In contrast to static circumstances engchi. tariff effects with the dynamic setting and adjustment process with a currency substitution. model. He finds that considering the intertemporal tradeoff, the imposition of tariff may be expansionary in the short run, but in the opposite direction in the long run. It is because national saving and current account surplus will take place and cause a gradually reversal. The two key assumptions of his work include the static and rational expectation. To explore more from static analyses with intertemporal optimization framework, Sen and Turnovsky (1989) build a one sector model to examine the tariff effects. They find that the introduction of tariff is contractionary both in the long run and in the short run. However, the decline of the capital stock is accompanied by a current account surplus, and therefore the effect 3.

(7) on saving is ambiguous. Considering that tariff may affect different parts of the economy, several studies are extended into a multi-sector framework (for example, Gavin 1991 and Turnovsky 1991). Brock and Turnovsky (1993) construct a two-sector framework and focus on the welfare analysis. They find that the imposition of tariff on both consumption and investment goods will raise welfare in the short run but the effect will decline over time. In recent years, most macroeconomic studies are done in dynamic general equilibrium framework, initiated by Obstfeld and Rogoff (1995) who emphasize micro-foundation, choice theoretic and market imperfect competition. This model was extended by Obstfeld and Rogoff (2000) to an explicitly stochastic environment. While Obstfeld and Rogoff mainly discuss two-. 政治大 Kollmann (2002) establishes a small open economy DSGE model where goods are tradable and 立. country frameworks, other studies use the small open economy framework. For example,. the law of one price does not hold to examine optimal monetary policy.. ‧ 國. 學. While the above studies do not consider tariff, Fender and Yip (2000) is the first one to. ‧. examine the macroeconomic effects of tariffs in short run and long run following the framework. y. Nat. of Obstfeld and Rogoff (1995). They draw the conclusion that domestic output and employment. er. io. sit. fall as a consequence of tariff imposition, but tariff’s effect on welfare is ambiguous. When they discuss the effects in the short run, they conclude that temporary tariff on import reduces both. al. n. v i n the domestic and foreign output. TheCanticipated tariff on export h e n g c h i U can only reduce foreign output. but have ambiguous effects on the domestic output. Hwang and Turnovsky (2012) discuss the imposition of tariff under nominal rigidity with a two-country DSGE framework to examine the effects of tariff under producer-currency pricing (PCP) and local-currency pricing (LCP). They find that the effects were irrelevant for the flexible pricing scheme, though the mechanism is different. Under nominal rigidity, nevertheless, the currency of pricing can crucially lead to different effects. While they emphasize the nominal rigidity where prices may fail to respond to the unanticipated imposition of tariff, they do not consider the anticipation of tariff. Considering their model which has only one asset (currency), Reitz and Slopek (2005) add an integrated bond market into their model. They want to discuss the situation where the current account can be 4.

(8) imbalanced in the short run, but the movement in the interest rate and capital. From the other side, instead of imposition of tariff, some other research focuses on the effects of eliminating trade barrier consistent with the global trend of integration. Konan and Maskus (2000) discuss the effects after the FTA is signed between Egypt and European Union. Because government revenue is reduced after the tariff is lowered, they have to raise the domestic tax rates to offset the tax revenue loss. However, they find that capital tax was more distorting than goods and service taxes and therefore will cause the appreciation and reduce the real returns. In the recent works with a New Keynesian framework, Ganelli and Tervala (2015) find that the unilateral trade liberalization is welfare decreasing. The benefit of the WTO trade. 政治大 While the previous works study the effects of imposing or eliminating the tariff, we are 立. agreement significantly depends on the within-country substitutability and Frisch elasticity.. interested in the effect of news announcement of tariff reduction which may affect people’s. ‧ 國. 學. expectation before the policy is implemented. This is related to the news shock in the literature.. ‧. The concept that the change of expectation can affect macro economy originates from Pigou. sit. y. Nat. (1926), who points out that the optimistic expectation for future may cause the recession when. io. er. the expectation is unrealized. The famous theory which is called the Pigou cycle unveils the new stream for the expectation theory. Led by Beaudry and Portier (2004, 2006 and 2007), the. al. n. v i n C hnews on the TFP improvement theory of business cycle driven by the becomes prevalent. The engchi U. forward-looking household has the incentive to react to the news before it is actually realized.. Beaudry and Portier (2006) provide an empirical study on the stock market to show that the news of TFP progress does affect the investment and consumption. They also provide a model which helps solve the “co-movement puzzle”.. 2. Their study helps identify which news can be. a factor to stimulate the business cycles, which is called “news driven business cycles”. Beaudry and Portier (2007) introduce a costly distribution system and consider the effects between multi-sectors to make the co-movement possible. Jaimoivich and Rebelo (2009) introduce. 2 In traditional RBC model, it is impossible to explain the simultaneous expansion of output, consumption and hours worked of labor after the news of innovation. This is so called “co-movement puzzle” 5.

(9) capacity utilization and utility function with no income effect in their model. Capacity utilization rate depends positively on consumption, and this makes labor demand shift in response to the change in consumption. The effect of labor demand and real wages introduce a strong substitution effect on leisure, and thus the co-movement becomes possible. Gou, Sirbu and Weder (2015) consider imperfect competition instead. When the producer expect the TFP progress, they will have the incentive to decrease price today. This leads to a fall in desired markup and an increment in real marginal cost, which let the labor demand increase and generate the co-movement. Stephane, Paul and Shen (2009) discuss the case where the news about the TFP is fully realized, unrealized, higher or lower than expectation. We would like to. 政治大 macroeconomics (NOEM) circumstance. 立. introduce same case into tariff news. And we also analyze in new open economy. While the previous studies on tariff all focus on the effects after the tariff policy is fully. ‧ 國. 學. known to the public, none has considered the effects of news on tariff, by treating the tariff news. ‧. and its aftermath similar to Stephane, Paul and Shen (2009).. sit. y. Nat. In this study, Section 1 is the introduction and Section 2 provides the literature review.. io. er. Section 3 outlines the model. In Section 4 and 5, we explain the specification of parameter values for simulations and the analyses.. n. al. Ch. engchi. 6. i n U. v.

(10) 3. Model 3.1 Goods market. We consider a small open economy with a representative household, firm, and central bank, following the specification of Kollmann (2002). There are two kinds of goods: a variety of intermediate goods and homogenous final goods. Final goods are produced using intermediate goods and are used for consumption and investment in the domestic market, which is nontradable. We assume that the intermediate goods are monopolistically competitive and are tradable. The types of intermediate goods s  0,1 , and they are tradable. The import, export. 政治大. and domestic use of intermediate goods are denoted with qtm  s  , qtx  s  and qtd  s . 立. respectively. Intermediate goods firm use capital and labor as inputs, which are immobile. ‧ 國. 學. internationally, to produce. The household owns all capital stock, which is rented to producers,. er. io. sit. y. Nat. 3.2 Firms. ‧. and supply labor. The labor and assets market are perfectly competitive.. 3.2.1 Final goods producer. n. al. Ch. engchi. i n U. v. The final goods producer uses the following production function . 1  1 1  1  1   Zt   d   Qtd     m   Qtm    ,  . (1). Z t is the final goods at period t, with  d ,  m  0 as the share of domestic and imported intermediate goods such that  d   m  1 . Qtd , Qtm are the aggregate indices of the domestic and imported intermediate goods, which are integrated as follows v. v 1  1  v 1 Q    qti  s  v ds  , v  1 , for i  d , m, x . 0  i t. 7.

(11) qtd  s  and qtm  s  are the domestic and imported intermediate goods of type s , respectively. Let Pt d  s  and Pt m  s  be the price of these two types of goods in the domestic currency set by producers. We assume that government imposes the tariff  t on import goods at period. t . Since the government collects tariff payments from importers, final goods producers will face the price Pt m , which includes the tariff. The effect of tariff will appear in the process of import firm’s price optimization. Export goods pricing follows the same mechanism, but with the tariff.  * on exports, imposed by the rest of the world. Under cost minimization of the final good producers, we obtain. 治政 P s  q s   Q 大. v.  Pt d  s   d q s    Qt , d  Pt  d t. Pt  m. ,.  P 1. 0. Nat. io. .  Pd  Q    t  Zt ,  Pt . t. m. s. v 1. . ds. . al. With. . Pt  . d.  Pm  Q    t  Zt  Pt  m t. n. d. P . d 1. t. . m. (2). Ch. . 1 m 1 1. P  t. m. engchi U. .. 1 v 1. .. (3). y. 1 v 1. Where Pt m embodies the tariff implicitly. We can also get. d t. .. sit. t. v 1. . er. 0. d. m t. ‧.  P  s  ds 1.  Pt. m. 學. Pt  d. ‧ 國. 立. With. v. m. t. m t. v ni. (2a). (3a). Pt d and Pt m are the price indices for each domestic and imported intermediate goods that are. sold in the domestic market. Pt is the price of final goods under perfect competition ( P  MC ).. 8.

(12) 3.2.2 Intermediate goods firms. We assume a representative intermediate firm which uses capital and labor to produce different kinds of intermediate goods s. The intermediate goods producer uses the following production function yt  s   At Kt  s  Lt  s  . 1. ,. 0   1 .. (4). where yt  s  is the intermediate firm’s output at period t . At is an exogenous productivity shock which is assumed to be identical for all domestic producers. Kt  s  and Lt  s . 政治大 represent the capital and labor used by home producers respectively. 立. ‧ 國. 學. and Wt be the rental rate of capital and wage. Under cost minimization we get. Let Rt. ‧. Lt  s  1   Rt ,  Kt  s   Wt. (5). sit. y. Nat. where the total cost TC is equal to the sum of wage and capital rental costs, TC  WL  RK .. 1.  Wt     1 . al. n. 1 MCt  At. er. io. Therefore, the marginal cost can derived as .  Rt    .  . Ch. engchi. i n U. v. (6). the production of intermediate goods are sold domestically and also exported. Therefore, we get the relation: yt  s   qtd  s   qtx  s  , where qtx  s  is for the export demand. We assume. qtx  s  follows the same form as the domestic and imported demand defined above, . v.  P x  s   Px  q  s    t x  Qtx , Qtx  X *  t *  ,  Pt   Pt  x t. η>0,. (7). Pt x  s  is the price that firm exports, denominated in the foreign currency. Pt x uses the same. 9.

(13) method to integrate and therefore Pt  x.  P  s  ds 1. 0. v -1. x. 1 v -1. t. and so does the aggregate foreign. demand for home goods Qtx . Pt * is the exogenous world price level and X * is the overall foreign demand which is assumed to be exogenous. Therefore, the profit of domestic intermediate good producer,  tdx , which can make profit from selling goods to domestic final goods firms and export to the foreign (producer), and also the costs of tariff payments, firm take tariff as a part of producing cost3. So we can then express the profit of domestic intermediate goods producer as  tdx , and of an intermediate goods. 政治大. importer, which buy intermediate goods from foreign country and sell it to domestic final goods firms for profit  tm :. 立. ‧ 國. 學.  tdx  Pt d  s  , Pt x  s  .   Pt d  s   MCt  qtd (s)    et Pt x  s   MCt  et t* Pt x  s   qtx (s) . . Nat. y. .   P x  s   v    t x  Qtx  ,   Pt    . ‧.   P d  s   v    Pt  s   MCt    t d  Qtd   et 1   t*  Pt x  s   MCt   Pt     d. sit er. io.  tm  Pt m  s     Pt m  s  - et Pt * -  t Pt m  s   qtm ( s) .  P s   ,   1    P a sl  e P     Q  iv C h   P  U n engchi. n. t. t. v. m. m. *. t. t t. m. (8). m t. t. where et. is the nominal exchange rate, defined as the domestic-currency price of one foreign. currency. Following the staggered pricing as set in Calvo (1983), each firm may reoptimize its price with the probability 1   in any given period, which implies its probability of keeping their price unchanged is  . Thus, the average duration of price change is 1    . When producer 1. sets its price, one should consider the intertemporal profit maximization regarding the price. 3 We assume tariff is imposed on import and export firms, so firms take tariff as a part of costs to set the optimal price. 10.

(14) which may be applied in the future periods. Firms are assumed to meet all demands at posted price. Pt i,t  is the optimal price that firms set at t   . Due to the continual inflation, if an intermediate good producer doesn’t receive a “price change signal”, he will set the price. Pt ,dt   Pt ,dt  in the period t   where  is the steady-state inflation rate. The domestic intermediate goods producer considers a profit maximization problem. d  x    tdx  P ,t  t , P  t s   E    t  ,t t Pt   0   . Max Ptd,t. . v. s.t.. v.  P d  s   Px  s  q  s   t  d  Qtd , qtx  s    t  x  Qtx  Pt    Pt   d t . 立. 政治大. ‧ 國. 學. where t ,t  is the pricing kernel, which is the intertemporal substitution between t and t   .. ‧. Consider a Home producer can undertake price discrimination between the domestic and foreign. y. Nat. markets under local-currency pricing (LCP), which means that the law of one price may fail. n. al. er. io. sit. such that Pt d  s   et Pt x  s  . Thus, we can obtain two optimal price setting as follows (sold in both the domestic market and foreign market). Ch. v  d     v    0   Et t ,t  MCt   d Pt ,t        v  1    1v  Et td,t    0  . . engchi. i n U. v. (9). .     * v   Et tx,t  MCt         0 v   Pt ,xt   ,     v  1     (* )1v  Et  tx,t  et  1   t*    0  . (10). where it ,t   t ,t   Pt Pt   Qti  Pt i  , i  d , x . * is the steady-state inflation rate of v. the world price level. imply the markup of the firms. Analogously, an importer who imports immediate goods from the foreign system will meet the same pricing structure, but with different . cost and discount factor for the future payoff Rt ,t  where Rt ,t  1 , Rt ,t    1  it*  . The  1. 11. 1.

(15) importer faces the profit maximization problem as. Max.   tm  Pt ,mt   Et  Rt ,t   et   0 . s.t..  Pt m  s   m q s    Qt  m  Pt  . . Ptm,t. .   . v. m t . An importer that gets to set a new price selects.           v Et  tm,t  Pt *    v   0 Pt ,mt    ,   1   t      v 1   1 v  m  ()  Et t ,t  e    0 t  . . . 立. tm,t   Rt ,t  Qtm  Pt m  .. ‧ 國. v. (11). 學. with. 政治大. Consider a share of d firms choose to keep their price unchanged, thus the aggregate price.   Pt x1*. io.  a1    P . 1v. n. x 1v. i 1v t ,t. , i  d, m. x 1v t ,t. .. l C hengchi U. y. P  t.   1    P . 1v. sit.   Pt i1. i 1v. er. Nat. P  t. ‧. indices follows. v ni. (12). 3.3 The representative household. The economy is populated by a large number of identical and infinitely lived households. For simplicity, population is normalized to unity. The utility of household comes from consumption and leisure, we simply assume a logarithmic utility function as U  C, L   ln  Ct   Lt  .. (13). According to the permanent-income theory of consumption, household has an infinite 12.

(16) planning horizon, so will consider his whole lifetime utility including the current and future periods. So the life time utility can be described as . E0  t ln  Ct   Lt . .. (14). t 0. Et represents the conditional rational expectation at period t . Ct and Lt denote consumption and labor effort of the household at time t .  which lies between zero to one is the subjective discount factor. The resources which household owns include the revenues from physical capital and wage profit from production and bonds. The household invests through buying capital, so the relation between investment and capital follows. Kt 1  Kt 1       Kt 1 , Kt   It .. (15). 政治大. Where K t and I t denote the household’s stock of private capital and gross investment.. 立. from. 1 2. firm’s. purchase. of. 學. generates. ‧ 國. 0    1 is the depreciation rate of capital. Function    describes the adjustment cost. capital,. which. is. defined. as.   Kt 1 , Kt     Kt 1  Kt  / Kt between t and t  1 .  is the adjustment cost that. y. Let B denotes the one-period nominal domestic bonds held by. sit. Nat. capital accumulation entails.. ‧. 2. al. er. io. households in terms of the domestic currency, B* be the foreign bonds in terms of foreign. v. n. currency. The household’s budget constraint can then be expressed as Bt 1  et B*t 1  Pt  Ct  It . Ch. i n U. engchi.  Bt 1  it 1   et B*t 1  it f1   Rt Kt  Wt Lt    tdx  s  ds  Tt . 1. 0. (16). Where it 1 and it f1 are the interest rate of the home and foreign bonds B and B* . Tt is the transfer payment. The representative household maximizes its expected lifetime utility (14) subject to (15) and (16) by choosing the variables. C , L , B t. t. , Bt*1 , Kt 1. . t 1. t 0. with the initial value B0 , B0* , K0 .. The optimality conditions with respect to the associated variables are. 13.

(17) Ct.   P   : 1  1  it  Et  t ,t 1  t   ,   Pt 1  . Lt. :. (17a). Wt  Ct , Pt. (17b).   P  e   B*t 1 : 1  (1  it f ) Et  t ,t 1  t  t 1   ,  Pt 1  et   . (17c). Rt 1     P 1    2,t 1 Kt 1 : 1  Et  t ,t 1  t  1  1,t 1     . (17d).     ,   . where (17a), (17c) and (17d) are the first-order equations with time preference discount factor. 政治大 Also, we assume that the interest 立 rate of foreign bonds follows a relationship with the. t ,t i   Ct Ct 1 . Eq. (17b) implies the marginal substitution between consumption and leisure.. ‧ 國. t. . *. * t. *.  Bt 1  *   Pt . . ‧. 1  i   1  i   f. 學. internationally prevailing interest rate. . (18). sit. y. Nat. Eq. (18) shows the relation between the interest rate of foreign bonds and the internationally. n. al. er. io. prevailing interest rate where it* is the world interest rate.   P xQ x Pt* is the steady state. i n U. v. real export.  characterizes the degree of capital mobility. The higher value of  implies. Ch. engchi. greater friction on the international capital market. When   0 , the international capital market is perfectly mobile. We assume that an imperfectly mobile capital market where   0 , for a stationary equilibrium exists as the model in Sims (2000).. 14.

(18) 3.4 Tariff and government. We assume that tariff is an exogenous variable which follows an AR(1) process. The tariff normally stays the same, but can change if there are some policy changes. In particular, in the recent trend of global integration, the formation of FTA may lower the tariff rates significantly. The policy shocks can include two parts: news shock and surprise shock. Following the specifications of news shock in Beaudry and Portier (2004) and Stephane, Paul and Shen (2009), tariff can then be expressed as. 政治大.  t   t 1  t   t ,. 立. (19)4. where t is the news announced before  period at time t , which means household. ‧ 國. 學. receives a news about “tariff will change at period t ” in date t   .  t is a surprise tariff shock without preannouncement. Both shocks are assumed to be white noise. Since the tariff. ‧. does not change often, it typically stays in the steady state where  t   t 1   .. y. Nat. io. sit. In the analyses in Section 4, we run simulations of various scenarios. When the government. n. al. er. announces the possibility of signing the FTA to lower the tariff, it may end up with the situation. i n U. v. in accordance with the news (realized), or the situation which is inconsistent with the news. Ch. engchi. because the negotiation fails (unrealized). Moreover, even if it comes to realize, the agreed tariff reductions can be higher or lower than expectation.5 Table 1 outlines the cases considered. The government sector takes tariff as the transfer payment, and does not execute active policies. Therefore, the budget constraint can be expressed as Tt   t  Pt m  Qm .. (20). This condition says that the tariff revenue is completely transferred back to the household.. 4. In the code for simulation, we assume that the tariff follows. problem of nonstationarity. Table 1 shows the various cases.. 5. 15.  t  0.01  0.99 t 1  t   t to avoid the.

(19) 3.5 Central bank. We assume that the central bank conducts a monetary rule following the Taylor rule (Taylor, 1993), which can be considered as:. ˆ d   Yˆ , it  i    t Y t. (21). ˆ d   d    , Yˆ  Y  Y Y .  d is the measure of domestic inflation rate. Y is where  t t t t t t the real GDP. i , Y denote the steady-state values of the interest rate and output.  and Y. 政治大. are the policy parameters that the central bank chooses.. 學. ‧ 國. 立 3.6 Market cleaning condition. ‧. sit. y. Nat. In the economy, there are final goods market, intermediate goods market, labor market,. io. er. rental capital market, bond market and currency market. When the economy attains the steady state, all markets should be cleared. Therefore, the market cleaning conditions require. n. al. Zt  Ct  It ,. Ch. engchi. yt  Qd  Q x ,. i n U. v. (22) (23). Lt   Lt ( s)ds, Kt   Kt  s  ds ,. (24). Bt  0 ,. (25). 1. 1. 0. 0. . Pt mQtm  et Pt xQtx  et Bt* 1  it f1   *t 1Bt*1. . (26). Eq. (22) shows that final goods are used for consumption and investment, and Eq. (23) shows the domestic and foreign uses of domestic intermediate goods. In equation (24), Lt and. K t are the labor and rental capital used for the production of intermediate goods Z t . Eq. (25) 16.

(20) assumes that the domestic bonds is not held by foreigners and are cleared domestically. Eq. (26) assures the current account balance where the net borrowing or saving in the foreign bonds is equal to the trade balance.. 3.7 Exogenous variables. In addition to the tariff shock, there are four other exogenous shocks which involve productivity, world Inflation, world interest rate and foreign demand for exports. All of them are. 政治大 0    立ln  A    ,. assumed to be exogenous variables and follow the AR(1) process as expressed below: ln  At   1   A. A. t 1. A t. A. 1. (27) (28). it*  1   i  i*   ii*t 1  * ti ,. 0  i  1. (29). 0   X 1. (30). . sit. Nat. y. ‧ 國. . ln X t*  1   X  X   X ln X t*1   tX .. ‧.  . 學. ln  *t   1   *  ln  *    *ln  *t 1   * t* , 0   *  1. n. al. er. io. where *t  Pt* Pt*1 .  n characterizes the persistence and  tn is independent white noise disturbance where n  A, * , i, X * .. Ch. engchi. i n U. v. 4. Calibrated parameters. To simulate the dynamics, we have to obtain the steady state as the initial value of all the endogenous variables. Following Kollmann (2001a), the measurement of parameter values for simulations are calibrated for the data of three small open economies, Japan, Germany and United Kingdom. We assume domestic and foreign interest rates to be identical, which means 17.

(21) that net assets position is equal to zero in steady state. And we assume r to be 0.01 roughly according to the average of long-term capital return, to let us capture the relation between capital return and interest rate: r  1  i   1  1  i f *  1 . Since we get  1  r   1 holds in steady state, the discount factor  is set at 1/1.01 . The share of using foreign intermediate goods  m is set at 0.3 which is consistent with the import/GDP ratio of German and U.K.67 Following Hopper and Marquez (1995), the price elasticities of aggregate exports and imports are set at     0.6 (see Eq. (2a), (7)), which are estimated by the median value of the three sample countries. The markup v / v  1 set by intermediate goods firms when setting the optimal price is set at 1.2, consistent with the three. 政治大 with the German and British data 立and suggests that the optimal duration of the reset price is. sample estimations by Martins et al. (1996). Lopez-Salido (2000) fits Calvo pricing equation. ‧ 國. 學. about 4 quarters which implies  is 0.75. The steady-state inflation rates are all set to be 1, which implies   d   x  m  *  1.. ‧. The capital mobility parameter  (see Eq. (18)) used by Kollmann is set at 0.0019,. sit. y. Nat. according to the estimation of Lane and Milesi-Ferretti (2011), based on the panel data of the. n. al. er. io. OECD countries. The technology parameter is set at   0.24 (see Eq. (4)), which is calibrated. i n U. v. for the steady-state income/GDP ratio to be about 60%. According to the aggregate data, the. Ch. engchi. quarterly depreciation rate  is set as 2.5%. The adjustment cost of capital accumulation  is set at 15. According to Kollmann, the setting of adjustment cost is set to match the fact that the standard deviation of the HP filtered log investment to be 3-4 times larger than that of the GDP in the sample counties. Using the data of three countries in 1973-1994 to estimate the total factor productivity (TFP) 8 , we specify the parameters associated with shocks  A  0.9,  A  0.01,  *  0.8 ,. 6. This ratio also roughly corresponds to the data in Taiwan. The import/GDP ratio  10% in Japan. The key results in Kollmann (2002) are obtained with this value. Since we want to discuss the effect of tariff, we choose the higher share of import goods to capture more significant effects. 8 The data correspond to those in Kollmann (2001a). 18 7.

(22)  *  0.005,  i  0.75 ,  i  0.004 . The steady-state tariff rate (  and  * ) for the import and export are specified as 0.17. The overall export demand X * is set at 0.27, to have the balanced current account in the steady state.9. 5. Analysis. Assuming the small open country is negotiating a FTA, to let the tariff between two. 政治大 both imports and exports. Generally, the decrement of tariff could proceed in two ways: (1) drop 立 economies decrease or even turn to zero. This decrement will definitely increase the amount of. step by step or (2) decrease immediately in an amount in agreement. To simplify the analysis,. ‧ 國. 學. we simply assume case 1 but only one step10. In the negotiation of FTA, both sides should take. ‧. benefits in the agreement, which means both import and export tariffs should decrease. But it. sit. y. Nat. has difficulties in simulating two effects simultaneously, so we analyze separately.. io. er. We try to capture the effects when a tariff news is announced. We assume there is a news on tariff decrease four periods before the negotiation is completed and the policy is implemented.. al. n. v i n C four In the following analyses, we assume First, we will compare the effects when the h ecases. ngchi U policy is preannounced and the effects if the policy comes as a surprise to see if there is an. advantage for preannouncement or not. Second, we will compare the scenario with different length of period (1, 4 and 8) preannouncement to analyze if the duration plays an important role or not. Third, we simulate the possible situation that the negotiation takes more time than expectation which is four periods here. Forth, negotiation turns to fail, thus the tariff decrement. 9. The calibrated value summarize into table 2 Considering the decreasing process of tariff in FTA will not turns down to zero immediately, the amount of decrement we assumed is 0.05, not full size 0.17. 19 10.

(23) is not realized due to some frictions during the negotiation which cause the tariff reduction is higher or lower than the news announced before.. 11. 5.1 Case 1: Tariff decrement with and without preannouncement. In this case we simulate an unanticipated decrease in tariff, which is referred to a surprise shock on tariff. We can tell from figure 1 that the decrease in tariff directly lowers the cost of import firms. Based on the price setting, the firm will decrease its price to reflect the cost. 政治大 investment. The lower import price reduces the aggregate price level which is composed of 立 reduction which leads to greater demand for imports and thereby the consumption and. both domestic and import goods prices. Because the lower price pushes the demand up, the. ‧ 國. 學. output of intermediate goods and the amount of domestic goods increase. Generally, the. ‧. decrement in tariff can strengthen the economy through the increase in the aggregate demand.. sit. y. Nat. However, things may go in the opposite way when we alert the tariff news ahead. We can. io. er. find that in figure 2, the output in the economy is lowered after the tariff news is announced, and realize the effects of tariff decrement since the policy is truly implement. The reason is that. al. n. v i n final goods producer will expect aCdecrease in import prices h e n g c h i U in the future, and hence will postpone their consumption in intermediate goods when the prices are actually lowered.. Therefore, the decrement of final goods output restrain the consumption and investment, as well as intermediate output, drop after tariff news is announced. This decrement of aggregate output also decreases the demand for labor. Note that the real exchange rate will increase (real depreciation) just at the moment the news is announced due to the current account deficit caused by the recession on output The depreciation makes the costs of import firms higher and leads to higher import prices. However, when the negotiation is completed and the tariff is actually. 11. The deviation from steady state is measure in quarterly effects. One should take summation of four period to see the effects of one year. 20.

(24) lowered, the positive effects of tariff decrement starts to dominate. The lower price for importers dominates the nominal depreciation, and therefore leads to lower import prices. The dynamics will be same as the previous case when the tariff is lowered. The consumption and output are higher. Thus, the preannouncement of the signing of FTA may have a negative effect in the short run by lowering the output before the negotiation is completed and people maintain their expectation for tariff reduction.. 5.2 Case 2: The time news on tariff reduction is released before implementation. 政治大. The following analyses are all base on tariff news preannouncement. In case 2, we simulate. 立. the different periods of the preannouncement of tariff reduction before the policy is implemented.. ‧ 國. 學. For period 4, 1 and 8, we can obtain figure 2, 3 and 4, respectively. Note that, despite the path of impulse response functions look alike, how long the reduction of the output lasts and the time. ‧. the economy recovers critically depend on the period between the announcement of tariff. y. Nat. io. sit. reduction and the policy is actually implemented. The result suggests that the earlier tariff. n. al. er. decrement news is released, the more impact it may have on the economy in the short run.. Ch. engchi. i n U. v. 5.3 Case 3: Policy delay. Another possible scenario is that it takes longer than expected to negotiate, so the actual implementation is postponed (we assume there is one year delay, in other words, 4 periods). The results are listed in figure 5. It is shown that this causes a kinked curve between the period the policy is expected to occur and it is actually implemented. This is just like the case unrealized what we going to analyze later. An interesting part is that, compared to the fully realized case, macroeconomic variables experience a sharper adjustment before the policy is implemented, 21.

(25) and the output will not start to increase until the fifth period, then gradually rise until the policy is realized. After the policy is revealed, the output will rise at a faster pace.. 5.4 Case 4: The tariff reduction is realized unexpectedly (news is unrealized or is realized at the size higher or lower than expectation). There can be various frictions during the negotiation which may delay the process or even result in the failure. Therefore, we simulate various scenarios: (1) Negotiation fails and thus the government cannot meet people’s expectation. The result is listed in figure 6. (2) Although the. 政治大 household’s expectation. The result is listed in figure 7. The results are quite intuitive. When the 立 contract is signed in the end, but the magnitude of tariff decrement is much lower than. news is just announced, the two scenarios follow the same path, but move differently when the. ‧ 國. 學. outcome of negotiation is realized. In the first case, the output does not rise above zero. Thus,. ‧. the economy suffers the negative impact of news, but does not get any benefits. In the second. sit. y. Nat. case, the economic expansion is weaker if the realized tariff reduction is smaller than expectation,. io. n. al. er. compared to that under the full realization.. 6. Conclusion. Ch. engchi. i n U. v. This paper analyzes the effect on macroeconomic variables when the government announces the news on a tariff decrement since the negotiation of FTA starts. With a smallopen-economy DSGE model, we examine various cases that the outcome of negotiation is consistent with people’s expectation. The news may affect people’s behavior by changing their expectation. We focus on the short run effects, to analyze the period between announcement and implement. The results show that the news does not generate positive effect, but negative in the 22.

(26) short run, because people tend to postpone their consumption and investment when the tariff reduction leads to lower import prices. The positive effect of tariff reduction will not occur until the tariff reduction is realized as expected. If the negotiation takes longer or the realized tariff reduction is smaller than expectation, the negative effect may last longer, followed by smaller or zero positive effects. The aggregate effects of tariff news in long run may affect the welfare. To evaluate its effects on the welfare, this model can be extended into one which focuses on welfare comparison. Another interesting issue remains for the future research is to examine the tariff news in an extended three country model to take into account the effects of the free trade agreement between two countries on the third country which is not included.. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. 23. i n U. v.

(27) Reference Brock, P. L., 1986. “Permanent and Temporary Tariff Reductions”. Unpublished Working Paper. Beaudry, P., Portier, F., 2006. “Stock prices, news, and economic fluctuations”. The American Economic Review, Vol.96, No. 4, pp.1293–1307. Beaudry, P., Portier, F., 2007. “When can changes in expectations cause business cycle fluctuations in neo-classical settings ? ”. Journal of Economic Theory, Vol.135, pp.458 – 477. Calvo, G., 1983. “Staggered prices in a utility maximizing framework”. Journal of Monetary Economics, Vol.12, pp.383–398. Clarida, R., Gali, J. and Gertler, M., 1999. “The science of monetary policy: a new Keynesian perspective”. Journal of Economic Literature, Vol. 37, pp.1661–1707. Eichengreen, B.J., 1981. “A dynamic model of tariffs, output, and employment under flexible exchange rate”. Journal of International Economics, Vol.11, pp.341–359. Fender, J. and Yip, C., 2000. “Tariffs and exchange rate dynamics redux”. Journal of International Money and Finance, Vol 19, pp.633–655. Fujiwara, I., Hirose, Y. and Shintani, M., 2011. “Can news be a major source of aggregate fluctuations? A Bayesian DSGE approach”. Journal of Money, Credit & Banking, Vol.43(1), pp.1(28). Gali, J., Monacelli, T., 2000. “Optimal monetary policy and exchange rate volatility in a small open economy”. Working Paper, Universitat Pompeu Fabra and Boston College.. 立. 政治大. ‧. ‧ 國. 學. sit. y. Nat. n. al. er. io. Ganelli, G., Tervala, J., 2015. “Value of WTO trade agreements in a New Keynesian Model”. International Monetary Fund WP/15/37. Guido L., 2011. “News and aggregate demand shocks”. Annual review of economic, pp.537–557. Guo, J.T., Sirbu, A,I. and Weder, M., 2015. “News about aggregate demand and the business cycle”. Journal of Monetary Economics, Vol.72, pp.83(14). Hooper, P., Marquez, J., 1995. “Exchange rates, prices, and external adjustment in the United States and Japan”. In: Kenen, P. (Ed.), Understanding Interdependence. Princeton University Press, Princeton, pp.107–168. Hwang, Y.N., Turnovsky, S.J., 2013. “Exchange rate pass-through and the effects of tariffs on. Ch. engchi. i n U. v. economic performance and welfare”. Journal of International Money and Finance. Jaimovich, N., Rebelo, S., 2009. “Can news about the future drive the business cycle?”. American Economic Review, Vol.99:4, pp.1097–1118. Konan, D.E., Maskus, K.E., 2000. “Joint trade liberalization and tax reform in a small open economy: the case of Egypt”. Journal of Development Economics, Vol.61, pp.365–392.. 24.

(28) Lane, P and Milesi, F.G., 2001. “Long term capital movements”. Working Paper, Trinity College Dublin and IMF. Lopez, S. D., 2000. Work in progress, Bank of Spain. Martins, J., Scarpetta, S., Pilat, D., 1996. “Mark-up pricing, market structure and the business cycle”. OECD Economic Studies, Vol 27, pp.71–105. Mundell, R., 1961. “Flexible exchange rates and employment policy”. Canadian Journal of Economics and Political Science, Vol.27, pp.509–517. Obstfeld, M., Rogoff, K., 1995. “Exchange rate dynamics redux”. Journal of Political Economy, Vol.103, pp.624-660. Obstfeld, M., Rogoff, K., 2000. “ New directions for stochastic open economy models”. Journal of International Economics, Vol.50, pp.117-153. Wang. P., 2012. “Understanding expectation-driven fluctuations: a labor-market approach”. Journal of Money, Credit and Banking, Vol. 44, No. 2-3 Pigou, A.C., (1926), “Industrial Fluctuations”, MacMillan, London. Reitz, S., Slopek, Q.D., 2005. “Macroeconomic effects of tariffs: Insights from a New Open Economy Macroeconomics Model”. Schweizerische Zeitschrift für Volkswirtschaft und Statistik, Vol. 141 (2), pp.285–311. Kollmann, R., 2001a. “The exchange rate in a dynamic-optimizing business cycle model with nominal rigidities: a quantitative investigation”. Journal of International Economics, Vol.55, pp.243–262. Kollmann, R., 2002. “Monetary policy rules in the open economy: effects on welfare and business cycles”. Journal of Monetary Economics, Vol.49, pp.989–1015. Rotemberg, J., Woodford, M., 1997. “An optimization-based econometric framework for the. 立. 政治大. ‧. ‧ 國. 學. sit. y. Nat. evaluation of monetary policy”. NBER Macroeconomics Annual, Vol.12, pp.297–346.. io. n. al. er. Sen, P., Turnovsky, S.J., 1989. “Tariffs, capital accumulation, and the current account in a small open economy”. International Economic Review, Vol.30, pp.811–831. Sims, C., 2000. “Second order accurate solution of discrete time dynamic equilibrium models”. Working paper, Economics Department, Princeton University. Stéphane, A., Paul G. and Shen G., 2013. “Nominal rigidities, monetary policy and pigou cycles “. The Economic Journal, Vol.123, pp.455–473. Taylor, J., 1993a. “Discretion versus policy rules in practice”. Carnegie-Rochester Conference Series on Public Policy 39, pp.195–214. Taylor, J., 1993b. “Macroeconomic Policy in a World Economy”. Norton, New York.. Ch. engchi. 25. i n U. v.

(29) Table. Table 1: The implication of various realization of tariff news (We only discuss the situation of enter FTA, so t is always greater than zero) Different combination of shocks. Implications. t  0. t  0. Anticipated tariff adjustment. t  0. t  0. Expectation full realized. t  0. 0   t  t. Expectation partially realized. t   0.  t  t . t  0. t  0. . 立. 政治Expectation 大 not realized at all. 學. Adjustment more than expectation. ‧. ‧ 國. io. sit. y. Nat. n. al. er. . Ch. engchi. 26. i n U. v.

(30) Parameter. Value. Description. r. 0.01. Capital return. . 1 1.01. Subjective discount factor. m. 0.3. Share of import goods use in final goods production. d. 1 m. Share of domestic goods use in final goods production. . 0.6. Price elasticities of aggregate import. . 0.6. Price elasticities of aggregate export. v. 6. Price-marginal markup composition. . 0.75. Probability of price resetting. . 0.0019. . . ‧ 國. Table 2: Calibrated Parameter. X*. 0.27. Export demand multiplier. . 0.17. Import tariff. *. 0.17. e. 1. A. 0.9. Lasting degree of last period of TFP. A. 0.01. Volatility of TFP. *. 0.8. Lasting degree of last period of world price. *. 0.005. Volatility of world price. i. 0.75. Lasting degree of last period of world interest rate. i. 0.004. Volatility of world interest rate. Inflation. 1. Price growth factor. 立Capital mobility. 學. Technology parameter for labor share. 0.025. Depreciation rate. 15. Capital adjustment cost. io. sit. Nat. y. ‧. 0.24. er. . 政治大. n. a lExport tariff v i n Ch n g c hratei U Nominaleexchange. 27.

(31) Figure12 Figure 1: Unanticipated tariff shock in period 4 (decrease by 0.05). 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. 12. Ch. engchi. i n U. v. The figures of impose response function here are measure in the level deviation from the steady state. To see the percentage change, the number should be multiplied by 100. 28.

(32) Figure 2: Anticipated tariff news shock (decrease 0.05 after 4 period). 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. 29. i n U. v.



(35) Figure 5: Anticipated tariff news shock (realize at one year postponed). 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. 32. i n U. v.

(36) Figure 6: Anticipated tariff news shock (unrealized). 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. 33. i n U. v.

(37) Figure 7: Anticipated tariff news shock lower expectation (news: decrease 0.05 after 4 period, real: 0.03). 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. 34. i n U. v.

(38) Appendix. Appendix A: Value function To let coding in dynare be possible, we have to transfer the form summation into the form of value. To associate with the conversion of real variables later, the computation combines both conversion of real variables and value function can refer as following.      E  d MC   t t ,t  t      0   Pt ,dt          1    1  Et td,t    0  . Multiplying both side with denominator, as substitute the td,t  into original form. . .  0.   E  1 . t. d t ,t . .  0.  . t. d t ,t . MCt .   d d   Qt   Pt     . ‧.     C  P Pt ,dt  0 1  Et    t  t   Ct   Pt .   E . 學.       立 1 . ‧ 國. Pt ,dt. 政治大. sit. n. al. er. io. Dividing both side with Ct Pt. Ch.   Q  d     Pt ,dt  0   1  Et    Pt       Ct  Pt   d t . engchi.  Qtd  d           E    t  C P   Pt   MCt       1   0  t  t  . Dividing both side with. y. Nat.         Ct  Pt  d d     Qt   Pt   MCt    0   Et     1   Ct   Pt  . P . d v. t.     Qtd  Pt d    1  P  0     Et   d    Ct  Pt   Pt    d t ,t. .  Qtd  Pt d             E  t  C P  Pd  MCt       1   0  t  t   t  35. i n U. v.

(39) Converting into real variables (dividing with Pt ).     Qtd Pt  Pt d    1  p  0     Et   d    Ct  Pt   Pt    d t ,t. .  Qtd  Pt d             E    d  MCt  t      1   0  Ct  Pt   Pt . Setting the value function, we can rename LHS as value function m(.) and RHS as n(t) ,. m(t)   0    . . 1 . .  Qtd   Pt d  Et    d  MCt   Ct  Pt    Pt  . d  1  Pt 1  +    MC     d  m(t  1) t   Pt   d  1  MCt +    t 1  m(t  1) . ‧ 國. 立. 政治大. 學.  Qd m(t)   t  Ct Pt  Qd m(t)   t  Ct Pt. .  Qtd  d  1 m(t)    mct +    t 1  m(t  1)  Ct  . y.  Q d P  P d  Et  t  t  t d   Ct  Pt   Pt  . sit. . 1 . ‧. . Nat. n(t)   0    . n. al. er. io.  Qd   P  P d  n(t)   t  +  1   t  t d1  n(t  1)  Ct   Pt 1  Pt    d    Qtd  t 1 1  n(t  1) n(t)    +      C   t   t 1 . Ch. engchi. i n U. v. where  td1  Pt d Pt d , then we get real optimal pricing of domestic intermediate goods in terms of value functions.     Qtd    1 d   mct +    t 1  m(t  1)        Ct  ptd,t       d    1    Qd   1   t 1   t n(t  1)    +        Ct   t 1   . Through this conversion Pt ,xt , Pt ,mt can also rewrite as following 36.

(40)        Qtx  * x     mct       t 1  o(t  1) Ct       x pt ,t   ,   x     -1    Q x q     t 1  * * - t t   1          *  j (t  1)     Ct    t 1 . . . . .   1  m * x  *   Qt      1  it 1   t 1    t 1  t (t  1)     m pt ,t   .  m * 1    1   Qt 1   t  1 * x    t 1     ( ) 1  it 1   t 1    s (t  1)   qt   t 1   . . . Where qt  et Pt* Pt ,  tx1  Pt x Pt x and  td1  Pt m Pt m. 立. 政治大. ‧. ‧ 國. 學. n. er. io. sit. y. Nat. al. Ch. engchi. 37. i n U. v.

(41) Appendix B: Converting into real variables. First of all, we have to summarize our model, and we can solve 40 endogenous variables with 40 equations. Solving. Zt , Qtm , Qtd , Qtx , Pt ,dt , Pt ,mt , Pt ,xt , Pt d , Pt m , Pt x , et , yt , At , Lt , Kt , Ct , I t , Rt , Wt , MCt , Tt , B, Bt* it , it f , it* , P, Pt* ,  t ,  t* ,s, t , m, n, o, j, xt* , td , tx , tm ..  P    1. 1  1  it  Et  t ,t 1  t     Pt 1   2.. (17a). Wt  Ct Pt. (17b). Rt 1     P 1    2,t 1 3. 1  Et  t ,t 1  t  1    1,t 1    . 政治大.   P  e   4. 1  (1  it f ) Et  t ,t 1  t  t 1   ,  Pt 1  et    5. Pt mQtm  et Pt xQtx  et Bt* 1  it f1   *t 1Bt*1.  Pm  8. Qtm   m  t   Pt . . (26). sit. (6). er. al. n. Lt  s  1   Rt  Kt  s   Wt. .  Rt     . io. 7..  Wt     1    . . (17c). y. Nat 1. 1 6. MCt  At. ‧. . (17d). 學. ‧ 國. 立.     ,   . Ch. engchi. i n U. v. (5). Zt. (2a).  Pt d  9. Q     Zt  Pt . (2a). . d t. d.  Pt x  10. Q  X  *   Pt  x t. . * t.  Qd 11. m(t)   t  Ct. (7).  1 d   mct +    t 1  m(t  1) . 38.

(42)  Qd 12. n(t)   t  Ct.   d   t 1 1  +        t 1.   n(t  1)  . . .  Qtx  * - x  13. o(t )    mct       t 1  o(t  1)  Ct . .  Qtx qt  * *  j ( t )  14.   1          Ct  15. s(t ) . .   x   t 1   t*1 .   j (t  1)  . * 1  Qtm 1   t     ()1 1  it*1   tx1   t 1  s(t  1) qt   t 1 . . 16. t (t )  Qtm     . .  1  i      t(t  1) x  t 1. 1 * t 1.     m(t )  17. ptd,t        1   n(t ) . 政治大. 立.     t (t )  18. ptm,t        1   s(t ) . * t 1. ‧ 國. 學.     o(t )  19. ptx,t        1   j (t ) .  1  d   Pt ,dt . Nat. 1v. 1v. 1v. 1v.  d  Pt m1 . 22.  Pt x . 1v.  d  Pt x1* .  1  d   Pt ,mt . 1v. 1. al. n. .  1  d   Pt ,xt . io. 21.  Pt m . 23. Pt   d  Pt d . 1v. y. 1v. sit.  d  Pt d1 . er. 1v. ‧. 20.  Pt d .   m  1   t  Pt. Ch.  e n g c h i. 1 m 1 1. i n U. v. (12) (12) (12) (3a). 24. Zt  Ct  It ,. (22). 25. Kt 1  Kt 1       Kt 1 , Kt   It. (15). 26. Bt  0 ,. (25). 27. yt  Qd  Q x. (23). 1  i   1  i   f. 28.. t. . *. * t.  Bt 1  *   Pt . . *. (18). . 29. Tt   t  Qm. (20). ˆ d   Yˆ 30. it  i    t Y t. (21) 39.

(43) 31.  t   t 1  t   t. (19). 32.  *   * 33. ln  At   1   A    Aln  At 1    tA ,. (27). 34. ln  *t   1   *  ln  *    *ln  *t 1   * t*. (28). 35. it*  1   i  i*   ii*t 1  * ti ,. (29).  . . . 36. ln X t*  1   X  X   X ln X t*1   tX. 38.  t  d. 39.  t  m. Pt d Pt d1. 政治大. Pt m m. Pt1 Pt x Pt x1. 立. 學. 40.  tx . Pt Pt 1. ‧ 國. 37.  t . (30). Equations (11.-19.) are already in terms of real variables.. ‧ y. Nat. io. al. er. Pt d Pm e P* Px , ptx  t * , ptm  t , qt  t Pt Pt Pt P. n. ptd . Ch. sit. And for the conversion into real variables, we let. engchi. i n U. v. Equations in terms of real variables can be expressed as following.   1   1. 1  (1  i) Et  t .t 1      t 1   2. w  C 3..   r  1    2,t 1     1  Et  t ,t 1  t 1    1  1,t     . 4..   q   1  (1  it f ) Et  t .t 1  t 1    qt   . 5.. ptmQtm  qt ptxQtx  qt bt* 1  it f1   *t 1bt*1. .  40.

(44) 1.  w  6. mc     1  7. rt . . r    .   mct  yt. 8. wt . Kt. 1     mct  yt Lt. 9. Qm   m  ptm  Zt . 10. Qtd   d  ptd  Zt . .  Qd 12. m(t)   t  Ct.  1 d   mct +    t 1  m(t  1) .  Qd 13. n(t)   t  Ct.     1   +         t 1  . d  t 1.   n(t  1)  . 學. ‧ 國. 立. 政治大. . . -   Qx  14. o(t )   t  mct    *   tx1  o(t  1)  Ct .   j (t  1)  . n. al. er. io. * 1  Qtm 1   t     ()1 1  it*1   tx1   t 1  s(t  1) qt   t 1 . . 17. t (t )  Qtm     . . . 1  it*1 . 1. C.  tx1h e  n*t 1 g  t (tch1) i .     m(t )  18. ptd,t        1   n(t )      t (t )  19. ptm,t        1   s(t )      o(t )  20. ptx,t        1   j (t ) .  . d 1 v t. 21. p. y. Nat. 16. s(t ) . .   x   t 1   t*1 . ‧. .  Qtx qt  * *  15. j (t )    1          Ct . 1 v.    d  ptd d   t . sit. 11. Qtx  X t*  ptx .  1  d   ptd,t . 1 v. 41. i n U. v.

(45) 1 v.  .     d  ptm m   t .  .    d  ptx x   t . m 1 v t. 22. p. 1 v. x 1 v t. 23. p. . 24. 1  . p . d 1 t. d. . m.  1  d   ptm,t . 1 v.  1  d   ptx,t . 1 v. . 1 m 1 1 t. 1    p . 25. Z  C I. 1   Kt  Kt 1  26. Kt   Kt 1 1     I t 2 Kt 1 2. 27. bt  0 28. yt  Qd  Q x. 1  i   1  i     b  29. f. * t. t. *. t 1. . *. 30. Tt   t  Qtm. ‧ 國. 學. d. 立. 政治大. 31. it  i   t   y Y t. ‧. 32.  t   t 1  t   t. io. er. 35. it*  1   i  i*   ii*t 1  * ti ,. al v i n C ln  X  h e ngchi U. n.  . 36. ln X t*  1   X  X   X 37.  td  38.  t  m. 39.  t  x. y. 34. ln  *t   1   *  ln  *    *ln  *t 1   * t*. sit. Nat. 33. ln  At   1   A    Aln  At 1    tA ,. * 1. t. X t. Pt d Pt d1 ptm t ptm1. ptx t ptx1. Where Pt and Pt * are no longer exist, and we separate (7.) before with production function yt  At Kt L1t into (7. and 8. ) here. Then we can solve 38 endogenous variables with 38. equations. 42.

(46) Appendix C: Solving the steady state In order describe the path of steady state, we have to obtain the initial value which we assume the economy is in steady state. First, we assume. q  1,    ,  *   * ,   *   d   m   x  1, i f  i  i* . 1. .  1, A  1, b  b*  0 ,. as what we discuss in chapter 3.9. Then we can derived the steady state as following 1. mc. . 1 . d. p . d 1 t. . m. . 1 m 1 1 t. p . 1. 1 1 1  d  v    m v 1    mc     qt    v  1     v  1  1.  v  1 m  qt   v 1 . 1.  v  d  mc   v 1 . 立. ‧ 國. ptx,t. ‧. v MC PX v mc  P   * * v  1 e 1    P v  1 eP* 1   * . ptx,t . v mc v  1 q 1   * . y. n. al. er. io. 3.. Nat. PX . sit. 2.. 1.  1   mc . 學. 1. v 1  1  m  v  qt   d  d   v 1  v  . 政治大. ptd,t. Ch. engchi. v Pd v MC   mc v 1 P v 1 v ptd,t  mc v 1. Pd . 4. Pt ,mt. Pm . v eP* Pm v 1 eP*   v  1 1    P v  1 1    P. ptm,t . v q v  1 1   . 43. i n U. v.

(47) 5. r     K t  2  K t 1  K t 1    K t  2  K t 1 2    rt 1  1      2   K  t  1   1  Et  t ,t 1  K K    t 1 t   1   K t 1   1  Et   r  1    1. .  r 1  . r. 1. .  1   . 6. w . r    . 1. ‧ 國 .   1   r. ‧. Nat. y. w  1    mc. 1 1. 學.      w  1     mc     r   . 立. 1 1. 政治大. io. sit. .    w  mc       r   1 . n. al. er. 1.  w  mc     1 . Ch. engchi. 44. i n U. v.

(48) 7. Z y  K .  L  1 y   K   L . 1. 1. r y   w. 1.  w K   r. 1.  w K   r 1.  w   r. 1.  1      . K 1.       1 . y. 1.       1  1.       1 . Q d  Q x . .  d  p d  Z  1   * p x     . . .  .      v mc     v  d    mc  Z      v  1 q  v 1         1 1    v mc     w     d  v  Z  w    mc  Z           r   1      v 1    v  1 q    1. 1. 1.  w Z    r. p x  ptx,t p d  ptd,t.  v  mc   v 1 . .  d . al. Ch. engchi. p m  ptm,t. 9. Qm , Qd , Q x. Qm   m  pm  Z . Qd   d  pd  Z . Qx  X *  px . 1.       1 . . n. p x , pd , pm.   v mc       v 1 q  . io.     1   . .   v mc       v 1 q  . sit. 1.  w 1    r. 1.       1 . Nat. 8..  w w   r.  v  mc     v 1  d. ‧. 1. Z. 1.       1 . 學. 1.  w Z  w   r. ‧ 國. 立. 政治大. y.       1 . er. 1.  w   r. . 45. i n U. v. .