1. Introduction As with the famous post-World War I European hyperinflations studied by Sargent (1993, chap. 3), Taiwan's was fueled by large budget deficits and exceedingly high rates of monetary expansion. Interestingly, though, prior to 1950 the deficits were not those of Taiwan per se but rather those run by the Nationalist regime in mainland China that used the proceeds from excess loans to public enterprises in Taiwan to help finance its war effort. By excluding loans to public enterprises from the official budgetary accounts, Taiwan's provincial government had the dubious distinction of reporting a (highly misleading) budget surplus in the midst of (see Li and Wu 1997). Consolidated fiscal data are not available until 1950, in which year the deficit amounted to 5.2% of Taiwan's GDP (see Table 1). Given that Taiwan is generally considered to have implemented a successful package in June 1949, such large deficits in 1950 seem to depart from the earlier European experiences where inflation was ended by a package that incorporated dramatic fiscal consolidation and movement away from deficit finance. Makinen and Woodward (1989, p. 103) conclude that effectiveness of the June 1949 reform package, in spite of continuing deficits and reliance on the printing press, implies that the Taiwanese public must have believed in a regime change. One explanation for such seemingly perverse behavior has been suggested by Li and Wu (1997), who argue that the June 1949 program could have balanced the budget had it not been for the subsequent retreat of the Nationalist central government to Taiwan in December 1949 that led to a dramatic increase in fiscal pressures. If these events were not foreseeable earlier in 1949, the June program's deficit reduction measures may have been credible ex ante even though later observation showed a widening, not a shrinking, budget deficit. A problem with this interpretation, however, is that by June 1949 the Nationalist regime had already been subjected to defeat after defeat on the mainland, culminating in the fall of Shanghai in May 1949. Prospects of Nationalist victory over the Communists were close to nonexistent; and, with the movement of Nationalist forces to Taiwan already accelerating in early 1949, the completion of this process later in the year was surely not a complete surprise. So, even allowing that the full budgetary consequences were subject to a lot of uncertainty, we still do not see how there could have been much justification for widespread optimism about the fiscal position in June 1949. Do these arguments lead us back to believing in a nonexistent regime change? No. The empirical work performed in this paper suggests that the inflation and money growth processes in Taiwan were, in fact, subject to a series of shocks from 1948 through 1950 and that these processes did not stabilize until well after the June 1949 policy announcements. This suggests that Makinen and Woodward's (1989) approach of dividing Taiwan's experience into a hyperinflation pre-June 1949 and a stabilization post-June 1949 is actually not supported by the available data. The evidence of structural breaks beginning in 1948 also raises concern with later work-such as Phylaktis and Taylor (1992)-that treats the pre-June 1949 period as a homogeneous sample.' While we do not pretend to resolve all the mysteries of the Taiwanese experience, we do believe that our allowance for multiple structural breaks highlights the fact that there was much more to the evolution of the inflation process than the June 1949 reform announcements. Indeed, we argue that other measures, both before and after June 1949, may well have been more important than the June 1949 package. 2. Origins of the Taiwanese Hyperinflation During World War II, Taiwan was ruled by Japan, and the Bank of Taiwan was forced to purchase a huge amount of national debt from Japan to help finance Japan's war effort. …