uodsuen pur $302 uonanpoid ypog 998) [Im wily Yoke ‘2pIS 1509 au UD “sioa4je $8039, vet Joyeo1 ase anuDAD! JeUTBseU UO 8192452 ,UMO, EY Burkes Deen aerate ses ea eect ea pone rene he he _ fe we ke ee Se pach ghey “o> re ee/.dre ip 0< 0 > Peery ane stu rayne aqp w Bussas29p 5] uakas [eUIBHWU s,4uII YOR Jey) WNSSE |B | (a = _s}uonoumy anata su uo.09 a AEG Afnnoadso sun wBlaoy pur auoy ay 4 (ip s¢ uusy auoy ayp Jo uonouny anuaaas ayp (1 ‘sunny ayp Jo suonsuny anu2sax 2y 0} Kpi9211P 08 pur suoN Supa ays dogs my 1-(z861) 29PUER pu AsOUIdg SuIMOII05 “1 =) eeu yy ‘sje pue Anunoo auroy s,uusy yore ui siexseW epnjour Kew Ax adAi 10 sox) 199109 'ss09 uodsuen £q papiarp 2q Aeui areduioo Aap Yoryy ut saxzew poruourfes ay souminsqns ayiod 2q 10u paw ing APw sionposd sus ‘ati “uy saqpo 2p Ypto UoRBadweo ur sioxzeUr Jo r9qWINY e UT SIIas 11 Yon “jonpoid offurs e senpord uuy ypeg “uEs10}-pue swoy ‘Suny om) axe aIayy, 41V9S 40 SIIWONODA DILVIS 31 13G0K 7 Pe reste per beet oe aiecanee 9 ayy swuasasd Z uon}ag “suon: jun uy Uno aneys 12qseU J2yy PuE ues ay StHewHas uorowosd se uoHD*0id Jo A1O}S o18eq ay 25FD YEE ‘suede sisuoayp apen pur wai jrarizerd so Smotn ayy 8195 Ajqensn py “SOMMEUAP UE soNeIS UIDMI>G UO! ayy 10 ‘sisoD Suisea:zep ue Sursea.9u uoomi0q ‘ssans pin zoded aap sy“ ‘edhy anno Buwsea, 27 30 So1urouoda apeos aqweukp ojdurexa 10} .synsax awies 2471 sanpo.d 0} sauwouoss 2189s Jo stwI04 afiqns asout 10) “ranamoy ‘ajqissod ose Surunpap & “91 “afe0s Jo sorouora 212836 29 pinom as% aye Aes asatig “21895 Jo SamwsOUOs9 Jo Pury atios st auioy $1910 yes om Surduinp yeroidioa1, wm a¥eBua ym sts 2 1y asm299q ‘AuisnpUt Ue urn ssouodxs pue si210duH yi0q 9g 01 $31 9POU! 2524) UF s1UDIPaIEUT o1seq Omi aie a1Oy uodxa se wonsaiod vodun jo mata s,uewssautsng aij) POU Jo Sseo St as3yp ey MOUS OF | 1aded SIAR Jo asodind ayy, ‘PUI UI axEY OF Woes UDLUSSaUISNG 2tp ey = Pouodxa 9q 01 vonanpo.d w Asevuaw2|duios st yonyn 2 Aews 1 “payodxa psmeo s9a9u ueD tonpoid iwloo 199}204 Jo pjiow & uy “a}qis 1 UI p3]OO4DS sysIWOUODS YOIYM jauindse ayy 0} ss2uianisensiad a[qequepun ue st aiay) ord 23 10921 0} Yous uIUIED Jo ‘yBneua Bui9s Jo “uonanpord 90969812 safeiuenpe Jo Joquunu ® axey “sund qusunBie ou awoy parsaioid & jo sFevueap uodxa uy suiny asouedey yo ssz20n5 ay) ur sweunsy jog 5 9]PIS JO SqtMoUodY pur sjodosyO jo souesalg BY} Ul UONNedWOD yeUOTeUIE}UT ‘UOHOWOIg WodxgY se uoNda}01g uodwy ~ (y36r ‘profxo) nus A104 st ansox puo ayy eg I 22, DENT PMD. o, ee i ie men ky tandong, (en rere ES ll \ ne ~ apeed, Ist ¥ cai Set Ng182 Import Protection as Export Promotion * P. Krugman aS ‘costs thus total costs for each frm will be: +C(E x) (11.6) ex, ain ‘ where we assume declining marginal cost of production: C”,C""<0. Notice may also be thied country markets to which presumably 4 <17; for the“foreign fimm's domestic nN How do these firms éompet i.e. it must choose about competition jn each market, Th is @ multimarket Cournot model, where the firms’ decision problems ae: max Tl = E Ray 3f)— Eh CE x) ans | PS 2 Ras: 2 apap —C%(E 2). (Ls) [ete 134 F eC ae The first-order conditions which determine equilibrium are ’ Fig. 11.1 Biyew eo iat pn =O f=1,....0 seine = 0 m1 compute the solution. Specifying this ‘no implications for the outcome; We then sum the chosen levels of deliveries to get total output and compute the implied marginal cost. This estimate of marginal ‘cost is then used for a second round and so on until convergence. The stages of n can be represented by the geometric apparatus presented. shows the ‘competition in a representative market for given estimates of marginal cost i tions of the domestic and foreign firm respeImport Protection as Export Promotion abd - (0) Fig. 11.3, RK Seelat yg eA lon tet + FRx OxF 3 3) and (11.4); whl iS compete, $0 that f the home firm will rise and total output of the foreign firm On one hand, the lower the firm's cures QO and MM, The, ‘equilibritim for the firm — conditional on the ‘other firm's ‘of marginal cost ~ is where MM and QQ tross. As drawn, Suppose now that the foreign fi cost, u*. This would imply a lef given 1, output of the home firm would The end result is that domestic marginal cost and vice vera, This takes us to the final step in determining equilibria then the dynamics is steeper than u* (qu), that his higher level. equil possible and even important that there may exist no stable ¢ except where one firm or the other ceases production. For the rest of this chapter, however, we will assume that there i 2 unique stable equim where both firms produce at positive levels Our next promotion device 3. EFFECTS OF PROTECTION Suppose that the home government excludes the foreign firm from some market previously open to it. This market might be the whole domestic market. or it nment-owned firms. For sim: ign product, although a quota ‘The effect under thisImport Protection as Export Promotion 1 py) Fig. 1.4 P. Krugman 187 cost cause FF to shift out, F°F* to shift in: x, rises, x] falls. Protecting the domestic firm in one market increases domesTic sales and Towers foreign sales inal markers This is the businessman's view, and it should be clear why it is confirmed, protectionas-promotion argument lies in more subtle forms of decressing cost. These are the dynamic economies of scale involved in the learning curve and in R & D. What | will do in the rest of this paper is show that these dynamic the protection as-promotion argument remains valid 4. MODELII: COMPETITION INR &D In this section 1 assume that marginal costs are constant. Firms can, however, production cost is independent of the level of output. but decteasing in the amount of investment each firm does in R & D. w= 4m) we), where ae aut ANG NE > ae at an?" av < © Profits of each firm are revenue, less production and transport costs. and also less R & D expense N= PRO) Laue) Lav ania) WM = CR Lee wy Lav 15) i 7 7188. Import Protection as Export Promotion repeatedly discussed: Spence (1981) is a recent example. I will opt for sim: plicity, and use the opendoop concept. This also has. the advantage of making iat investment in R & D has an effect 0: of increasing The effect of reserving some market for the domestic fi At given V and N* domestic output rises and foreign output curve shifts o counterpart shifts Shifts down: N rises, N" falls. Reduced m firm and higher marginal production costs for the foreign fie me domestic sales in al markets, tection, by increasing the home firm’s sales and sn competitor, increases the incentive for domestic R&D at foreign expense. This in turn translates into a shift in relative P. Krugman 9 yi ” < ff va a SeaImport Protection as Export Promotion jodel we comsier yet another form of eetnomis of see version there se neither satc economies of see or exp vest R& B inst a RS aees ret higher output turn out to yield results very simi ae: but now they compete verte aswell 3 pce In each mare the revenues of the two firms are: ' 2 i 18) 19) ne; otherwise they have R where xy. x7 now represent the same properties wehave been assuming all along, On the cost side, each firm faces constant transport costs market. At a point in time, production cos marginal costs 1,51 output. Let = Ex, ” to each are characterized by constant ‘these marginal costs are, however, dependent on previous (11.20) ‘The learning curve assumption is that marginal costs are a decreasing func- tion of cumulative output to date: make the ex! over a fixed obje: What does the optimum solution look ike? By selling another u 4, the firm gains two things: the direct marginal revenue, and the 191 der ion at a point aR, T ae en. Muff Boe 23) must also be ‘economic implication of wat on the basis of a constant shad ms ‘Again we can imagine an iterative procedure for calculating €q) ccan make a guess at the firms" terminal marginal costs sy, uz: find the cumu| from these guesses, and the corresponding 1 going into de be the same as in 0% model. Each seeps ‘The effect of protection is now exactly parallel to its effect in the case ot static seale economies. Excluding the foreign firm from some market the cumulative output of the domestic firm and reduces the cumulative ‘of the foreign firm for given uy, wp. The result is that zy (7 in means that x; rises and x? falls in all ly protected oF not. 6. SUMMARY AND CONCLUSIONS ‘The idea that a protected domestic: marke suceessiul exporting is one of those heterodox arguments, common in discussions of int national trade, which are incomprehé in terms of standard models ¥ ‘seem persuasive to practical men. This paper has developed some simple models ‘which make sense of the argument for protectiowas-promotion. To get hetero: ox conclusions one needs heterodox assumptions: these models assume oligo ead of perfect competition, decreasing costs instead of constant return. ingly, however, the economies of scale need not be simple static pro: ‘economies but can take fairly subtle dynamic forms192 Import Protection as Export Promotion that the US should prot the paper contains no welfare is extremely complex. We are comparing jons in any case; and if markets like the ones portrayed here I not be able to use the standard tools of consumer and second are prevalent, producer surplus, the difference between the conclusions of this paper and standard ‘based on differences in assumptions: which view is more ‘an empirical matter. Showing that heterodox ideas are self: w about here can be modelled. And itis impor at we try. It ‘may’ be that free trade and laissez-faire are good policies, and that most intr ventionist suggestions are selfserving fallacious, or both. But the argument of ade theorists will remain unpersuasive unless their models begin to contain ast some of the features 0% the world which practical men accuse them of P. Krugman 193 REFERENCES Brander, J. A, (1981), “Intra-industry Trade in Identical Commodities’, Journal ‘of International Economics, Brander,J. A. and Krugman, P. International Trade’, mimeo, M.LT. Spence, A. M. (1981), "The Learning Curve and Com}