摘要：The focus of this paper is on the relationship between public debt and economic growth in the ten Canadian provinces using data from 1990 to 2012. The study co n- cludes that the direction of causality is opposite to that claimed by Reinhart and Rogoff (2010) in their cross - country study: slow economic growth causes high public debt in Canada's provinces, not the other way round. In addition, we find no evidence sug ges t- ing a threshold in the relationship between public debt and GDP growth for high debt - level provinces. var currentpos,timer; function initialize() { timer=setInterval("scrollwindow()",10);} function sc(){clearInterval(timer); }function scrollwindow() { currentpos=document.body.scrollTop; window.scroll(0,++currentpos); if (currentpos != document.body.scrollTop) sc();} document.onmousedown=scdocument.ondblclick=initialize 40 Myatt and Liu Debt and growth in the Canadian Provinces Reproduced with permission of the copyright holder. Further reproduction prohibited. sality. For example, Bivens& Irons (2010) pointed out that RR offered no evidence that the correlation ran from high debt to low growth rather than the other way around, and other evi- dence suggested that the latter was more likely. The question of causality was tak- en up by Dube (2013). He tested the direction of causality by regressing past, present, and future GDPgrowth rates on the current debt-to-GDP ratio. He found a clear negative correlation between the debt ratio and contem- poraneous growth, and between the debt ratio and past growth (a 5-year past average), but almo st no relation- ship between the debt ratio and fu- ture growth (a 5-year forward aver- age). This is consistent with reverse causality (low growth causing a high debt-to-GDP ratio). After identifying this issue, Dube (2013) used several methods to address the endogeneity issue on debt caused by reverse cau- sality. He concluded that the relation- ship between the debt-to-GDP ratio and growth is flat for debt ratios ex- ceeding 30 percent. What we learn from this is the key importance of taking into account causality and endogeneity. This be- comes the focus of our own study in the context of Canada's provinces. Empirical Results Our data cover the ten Canadian pro v- inces (British Colombia, Alberta, Saskatchewan, Manitoba, Ontario, Quebec, New Brunswick, Prince Ed- ward Island, Nova Scotia, and New- foundland& Labrador) from 1990 to 2012. 4 Figure 1 shows the debt-to-GDP ratios in the five western provinces (British Columbia to Ontario), while Figure 2 shows the five eastern pro v- inces (Quebec to Newfoundland & La- brador), from 1990 to 2012. The overall picture is one of increasing indebted- ness in the early 1990s, followed by gradual decline until the early to mid- 2000s. Since then most provinces show quite a rapid increase in debt-to- GDP ratios again. Notable exceptions are Newfoundland-Labrador, Nova Scotia, and Saskatchewan.Figure 3 shows the average debt-to-GDP ratio by province for the same period.These data show that no province had a debt-to-GDP ratio over 60 percent at any time. This suggests that we need to modify the debt categories used by RR. If we blindly applied RR's defin i- tions to the provinces, none of the provinces would fall into the "high debt" (debt-to-GDP ratio between 60% and 90%) or "very high debt" (> 90%) categories. This seems inappropriate for several reasons. First, there are good reasons to believe that a sub- national government in a federal state cannot sustain levels of debt as high as the federal government. Foremost among these reasons is that federal debt is ultimately backed up by the money supply, whereas provincial debt is not; furthermore, in the Cana- dian context, the provinces have (by a mixture of tradition and law) less tax room than the Federal government. Thus, a debt to GDP ratio that may be only moderate for the Federal gov- ernment may be high for a Canadian
