back start next
[start] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [ 100 ] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] [133] [134] [135] [136] [137] [138] [139] [140] [141] [142] [143] [144] [145] [146] [147] [148] [149] [150] [151] [152] [153] [154] [155] [156] [157] [158] [159] [160] [161] [162] [163] [164] [165] [166] [167] [168] [169] [170] [171] [172] [173] [174] [175] [176] [177] [178] [179] [180] [181] [182] [183]
100 Supply Shocks Ball and Mankiw (1995) use the observation that large disturbances are more Ukely than small disturbances to cause firms to adjust their prices to develop and test a theory of supply shocks arising from costs of price adjustment. To understand their idea, consider an economy where there are costs of price adjustment, where firms are subject to relative cost shocks, and where aggregate demand is constant. In such a setting, only firms subject to unusually large relative cost shocks, either positive or negative, adjust their prices. The average relative cost shock is zero by definition. But if the shocks distribution across firms is skewed, there may be more large positive shocks than large negative ones, or vice versa. If, for example, the distribution is positively skewed, as in Figure 6.6, more firms raise their prices than lower them. Thus the average price level rises, and output falls. In a period when the distribution is negatively skewed, on the other hand, the price level falls and output rises. Thus changes in the skewness of the distribution of relative cost shocks act as aggregate supply shocks. To test this idea. Ball and Mankiw proxy the distribution of relative cost shocks with the distribution of relative price movements in disaggregated U.S. Producer Price Index data. They consider various measures of the asymmetry of this distribution. Their simplest measure is a weighted average of relative price movements that are greater in absolute value than some cutoff. That is, their measure is St = r = ~oc rft{r)dr + rft(r)dr, (6.93) where ft(r) is the density function of relative price changes in period t and Y is the cutoff. Note that if X = 0, 5 is simply the average relative price movement, which is zero by definition. Loosely speaking, 5 is the change in the aggregate price level caused by the price changes of industries whose relative prices change by more thanX. Thus if price adjustment is described The lack of a discernible link between < and , however, is a puzzle not only for the Lucas model but also for models based on small frictions: an increase in the variability of shocks should make hrms change their prices more often, and should therefore reduce the real impact of a change in aggregate demand. xariability terms, in contrast, play little role. The null hypothesis that the coefficients on both ax and are zero cannot be rejected at any reasonable confidence level, and the point estimates imply that reasonable changes in ax have quantitatively small effects on ; for example, a change in ax from 0.05 to 0.10 changes by only 0.04. Thus the results appear to favor the new Keynesian view over the Lucas model.
Density Firms do not change prices Relative cost shock FIGURE 6.6 prices The impact of the skewness of relative cost shocks on average precisely by the view shown in Figure 6.6, 5 is the appropriate measure of supply shocks. Ball and Mankiw compute 5 annually for the United States for the period 19481989. They focus on the case ot X = 10%, although their findings are robust to reasonable changes in this value. The resulting series for 5 exhibits large fluctuations over time. Some of the extreme observations correspond to wellknown supply shocks; for example, 5 is large and positive 1973 and 1979, years when there were large increases in the relative price of oil. But other cases do not correspond to previously identified supph shocks; for example, 5 is large and negative in 1952 and 1953. Ball and Mankiw test their theory in three ways. The first is to investigate whether 5 is associated with inflation. Regressing inflation on a constant, lagged inflation, and 5 yields TTt = 0.015 + 0.25277fi + 733St. (0.004) (0.082) (0.80) (6.94) Fr2 R = 0.765, D.W. = 2.01, s.e.e. = 0.023. Thus there is an overwhelmingly statistically significant relationship between the distribution of relative price movements and overaU inflation. The second test is to ask whether the skewness of relative price shocks affects the inflationunemployment tradeoff. In other words. Ball and Mankiw ask whether 5 shifts the PhQlips curve. Adding detrended unemployment, U, to the previous regression yields
= 0.015 + 0.283T7fi  1.15(7f + 6.84Sf, (0.004) (0.077) (0.44) (0.77) = 0.796, D.W. = 2.12, s.e.e. = 0.021. (6.95) Thus this change does not noticeably weaken the results. Ball and Mankiws final test is the most demanding of their theory. The conventional view is that the prices of oil and other raw materials are more flexible than other prices; as a result, disturbances to these goods relative prices affect the price level for a given level of output. In Ball and Mankiws model, in contrast, large shocks to the prices of these goods are supply shocks simply because they are large; which specific goods are involved is irrelevant. To test between these two views. Ball and Mankiw add the change in the relative price of raw materials to their equation. Their model predicts that this relativeprice measure should provide no information about aggregate supply once the overall skewness of relative price movements is accounted for; the conventional view predicts the reverse. The results are* = 0.018 I 0.257T7f 1  1.35[/, + 6.845f + 0.133Rf (0.004) (0.096) (0.43) (0.77) (0.136) (6.96) R = 0.812, D.W. = 2.25, s.e.e. = 0.020, where R is the change in the relative price of raw materials. The coefficient on R is small and insignificant; in addition, it is only about onefifth as large as it is in a regression that does not include the skewness measure. The coefficient on 5, in contrast, is little changed by the inclusion of P. In sum, the menucost models predictions about aggregate supply appear to be strongly supported. Microeconomic Evidence on Price Adjustment The central assumption of the analysis of this part of the chapter is that there is some kind of barrier to complete price adjustment at the level of individual firms. It is therefore natural to investigate pricing policies at the microeconomic level. By doing so, one can hope to learn both whether there are barriers to price adjustment, and if so, what form they take. Prominent examples of such studies include Carlton (1986), Cecchetti (1986), Lach and Tsiddon (1992), Blinder (1994), and Kashyap (1995). There are two general themes to the results. First, infrequent price adjustment is Ttie regression also includes a dummy variable for the Nixon wage and price controls; the dummy is not important for the results, however.
[start] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [ 100 ] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] [133] [134] [135] [136] [137] [138] [139] [140] [141] [142] [143] [144] [145] [146] [147] [148] [149] [150] [151] [152] [153] [154] [155] [156] [157] [158] [159] [160] [161] [162] [163] [164] [165] [166] [167] [168] [169] [170] [171] [172] [173] [174] [175] [176] [177] [178] [179] [180] [181] [182] [183]
