If no other disturbances occur, however, these payments and trade deficits will be eliminated over time without any major real consequence. The initial adjustment of the balance of payments represents mostly a portfolio-composition move, operating through the capital account, while the subsequent adjustment represents mostly a move to reduce the portfolio size, operating through the trade balance. This is the reduced-form equation for output, given by a weighted average of long-run deviations of agricultural output and long-run deviations of business sector output . In the long run, supply matches demand in both sectors. In the short run, deviations from the long-run equilibrium values are due to producers’ decisions in the agricultural sector and in the non-farm sector they are due to demand pressures. Appreciation of the exchange rate by increasing money growth raises inflationary expectations, reduces the real rate of interest, and therefore expands aggregate demand in both sectors. Equation shows that a real expansion in the money stock raises output in the non-farm sector and raises demand in the farm sector. According to 0, output will rise proportionally in the farm sector also. However, real depreciation has the opposite effect on agricultural supply, whose elasticity is given by ~. Hence, when the economy is away from the stationary equilibrium,growing blueberries in containers a change in relative prices has an ambiguous effect on the supply of agricultural output. While an improvement in the domestic relative agricultural price will directly induce higher output, it could also indirectly lead to its reduction through the effect on demand.
The total effect of a change in relative prices on the supply of agricultural output will depend on the relative size of these effects. The interval period chosen for the estimates will be of quarterly data from 1972: 1 to 1988:3. The endogenous variables will be LM2 , LER ., LNP , and LFP . The exogenous variables will be IR , LAIP , LAS , and LGEA . As far as the lag structure of the model is concerned, the evidence from Table 1 shows that IR and LGEA should enter the model with current values, whereas LAIP and LAS should not. This result is sensible because both the money supply and the exchange rate are expected to react quickly to changes in the interest rate brought about by changes in the domestic and foreign capital markets. Government expenditure, as a policy variable, is expected to counter expected movements in prices, not only to accommodate them. Finally, both foreign prices and farm stocks enter with lagged values since they tend to anticipate future values of actual market prices. Therefore, IR and LGEA will enter the model with their current and lagged values while LAIP and LAS will only enter with lagged values. Several different models were estimated, all with a fourth-order lag structure , with different sets of exogenous variables. The results of the estimation of every single model are not reported here. Here, we report only the MA representations of the full model. The impulse response functions of the full model are shown in Table 2. The model is estimated with the four exogenous variables, IR, LAIP, LAS, and LGEA. Here, we focus on only the MA representations obtained from the orthogonalized model . In most of the cases, in fact, the innovation correlation matrix remains almost unaltered and nearly diagonal. Money innovations have a persistent decreasing effect on manufacturing prices, more than proportional to the increase in money supply in the long run.
Money supply expands in the short run but not in the long run. Thus, money does not appear to be neutral, both in the short and in the long run with respect to manufacturing prices, but is neutral with respect to farm prices. Farm prices increase faster than manufacturing prices in the short run, which means that farm prices appear to be less sticky than non-farm prices. Finally, following a monetary unanticipated expansion, the exchange rate depreciates in the short run but appreciates in the long run. The effects of a monetary shock in the conditional model are quite different from those obtained with the unconditional model. The depreciation of the exchange rate lasts only a few quarters, whereas its long-run appreciation is more than proportional to the initial increase in money supply. However, as the theoretical model would predict, after an increase in money supply we have an initial depreciation in the exchange rate is experienced which is followed by an appreciation in the long run. Conversely, the effect of a monetary shock on prices is ambiguous. Farm prices increase in the medium term but are unchanged in the long run; manufacturing prices steadily decrease. There are two possible explanations for this result. The first is that the initial advantage in relative prices over the foreign sector is turned in favor of the farm sector when there is government intervention in agriculture. The second, more plausible, explanation is that expansions in money supply are subsumed by the interest rate and the fall in domestic manufacturing prices correctly “anticipates” the exchange rate appreciation that will come in the long run. In any case, this result seems to contradict the hypothesis that monetary shocks put the farm sector in a cost-price squeeze. Exchange rate innovations have the expected effect. Manufacturing prices increase initially but then decrease in the long run. It seems that non-farm prices are stickier, although this result is more attributable to the slow pass-through of exchange rate changes to prices.As in the unconditional model, farm prices respond faster but decrease less in the long run.
Money supply does change in the short run. This result is consistent with the view that the effect of an exchange rate shock is reflected in the interest rate only after a while, while it contracts in the long run. Hence, as the model would predict, exchange rate appreciation has a deflationary effect which is dampened by government intervention for farm prices. Also, money supply increases in the short run but not in the long run. Finally, the trade balance indeed appears to adjust slowly to movements in the exchange rate. The effects of manufacturing price innovations on farm prices and on the exchange rate are also as expected. However, the final money supply increase tends to be more similar to the long-run increase in manufacturing prices. This confirms the proposition of a neutral effect of price movements on money. The negative effect on the exchange rate is strong and persistent, whereas the effect on farm prices is positive although small. Farm price innovations have very unexpected effects. The resulting contractionary long-run decrease in money supply is more than proportional to the increase in farm prices so that farm prices are not neutral with respect to money. The decrease in manufacturing prices is, in any case, lower. In sum, price shocks are not neutral , but regardless the agricultural sector does not seem to be pushed into a cost-price squeeze. Again, this is in line with the theoretical model of section 2 where movements in farm prices are fully reflected in movements in money and the exchange rate. However, the effects of the two prices are different, which seems to confirm that farm prices are supply driven and manufacturing prices are more demand driven. In conclusion,square pots the analysis of the impulse response functions obtained with the conditional VEe model shows that, by explicitly accounting for the influence of relevant exogenous variables, some of the most important results already acknowledged in past studies are confirmed while some others are not. First, the neutrality proposition seems generally not to hold, particularly in the long run. Second, farm prices react more quickly to changes in money and the exchange rate; and, in the long run, their change is more than proportional to the change in manufacturing prices. Third, the backward effect from prices to money is less significant for manufacturing prices than for agricultural prices. Fourth, following a monetary shock, neither the exchange rate nor non-farm prices overshoot their long-run value.
Conversely, farm prices do overshoot their long-run value. These issues are explored more thoroughly in the next section through a set of specific hypotheses. Both the forward effect and the backward effect hypotheses can be simply tested as linear restrictions on the block coefficients in the conditional VEC model. The forward effect hypothesis, i.e., the hypothesis that changes in money and in the exchange rates result in changes in prices, can be examined by testing the null hypothesis that lags in either money or the exchange rate do not affect prices. Similarly, the backward effect hypothesis, i.e., the hypothesis that changes in prices result in changes in money and/or the exchange rate can be investigated by testing the null hypothesis that neither price effects money or the exchange rate. The first hypothesis encompasses a test of the forward linkages, whereas the second hypothesis encompasses a test of the backward linkages.? We have computed the tests of the feedback and the feed forward hypotheses for two VEC models. The first is the VEC model with no exogenous variable included. The second is the VEC model with four exogenous variables . The tests for the first model are reported in Table 3. All of the dynamic interactions are analyzed in a “closed” system, with no intervening effect from the outside. The hypothesis that prices do not affect money or the exchange rate cannot be rejected at any reasonable significance level. This implies that we are not able to reject the null hypothesis that there is no feedback from either price to money or the exchange rate. The hypothesis that money and the exchange rate do not affect prices, on the other hand, can be rejected. This implies that there is feedback from money or the exchange rate to prices. Table 4 reports the results of the tests for the conditional VEC model. Here, the evidence does support both hypotheses. Both the hypothesis that lags in prices do not affect money and the exchange rate and the opposite hypothesis, that lags in money and in the exchange rate do not affect prices, can be rejected at any acceptable level. In summary, while the evidence from these two sets of tests is mixed, it confirms that a proper test of the feedback hypotheses conducted with the fully conditional model would give different but more trustworthy results. In the unconditional case, we are expectedly led to the conclusion that money affects prices but not vice versa. In the conditional case, the conclusion is that, overall, the linkages are evident in both directions. In moving from the estimated reduced form of the model to the structural form, one obvious issue that arises is simultaneity. If all the variables entered the right-hand side of the model in lagged form, then the form of the exogeneity tests reported above are appropriate for testing those restrictions in the reduced form needed to recover the structural form. However, because the matrix of the contemporaneous coefficients is not necessarily null, we must proceed by testing restrictions that enable us to recover the structural coefficient matrix. In Tables 5 to 8, the four equations in our model are reported. In column of each table, the coefficients are given for the reduced-form equation estimated with the conditional VEC model with four exogenous variables. If there were not simultaneity among the four endogenous variables, then each equation could be interpreted as a structural equation. If we interpret the money equation as a structural equation, then it is not expected that the sum of the coefficients of non-farm prices will be positive. By allowing for simultaneity, we can examine the robustness of this result. We can then re-estimate the same equation with contemporaneous values of the other endogenous variables and investigate the sum of the coefficients. If contemporaneous values are indeed irrelevant, then the sum of coefficients at the bottom of columns and should be equal. Only for non-farm prices does this not seem to be the case. Non-farm prices appear to be simultaneously determined with money, at least to a degree.