More states are slowing even as the Nation continues to expand. The states that contracted include Wisconsin and Kansas, states pursuing a contractionary fiscal policy. Wisconsin’s level of activity lags that predicted by historical correlations.
The Geographic Dispersion of Growth. A diffusion index measures the proportion of individual entities expanding or contracting; a reading greater than 0 indicates a majority (of states, in this case) are expanding.
Figure 1: One month diffusion index (blue), and three month diffusion index (red). NBER defined recession dates shaded gray. Source: Philadelphia Fed, NBER.
Faster Growing States, Slower Growing States. A large majority of states are still expanding. This is manifested in a still-rising national coincident index.
Figure 2: Log coincident indices for Minnesota (blue), Wisconsin (red), Kansas (green), California (teal), US (black), all normalized to 2011M01=0. NBER defined recession dates shaded gray. Source: Philadelphia Fed, NBER, author’s calculations.
One of the few (non-energy dependent) states declining on a three month basis (see map) is Wisconsin. Fortunately, Kansas’s poor performance casts in a favorable light Wisconsin’s otherwise lackluster performance.
How Badly Is Wisconsin Lagging? While Wisconsin economic activity is declining, it might be the case that this is because national economic activity is decelerating, and Wisconsin follows the Nation. This hypothesis can be assessed by measuring the historical correlation between Wisconsin and the US up to 2009M06 using an error correction model, and then conducting out-of-sample forecasts (an ex post historical simulation). The forecast plus 90% forecast interval is shown in Figure 3. The methodology is laid out in this post (except a 2 lag specification is used, and is estimated over a sample starting in 1990).
Figure 3: Wisconsin coincident index (bold red), and dynamic out-of-sample forecast (purple), and 90% band (gray lines). Green denotes in-sample period. Source: Philadelphia Fed, NBER, author’s calculations.
The counterfactual takes account of the fact that Wisconsin economic activity has over the long term (1990-2009M06) lagged US economic activity; that is, a one percent increase in national economic activity is associated with a 0.73 percent increase in Wisconsin. For more details on inferring long run relations from error correction models, see here.
I have not assumed weak exogeneity of US economic activity for Wisconsin. Doing so — which seems plausible –yields the following figure.
Figure 4: Wisconsin coincident index (bold red), and dynamic out-of-sample forecast (purple), and 90% band (gray lines). Green denotes in-sample period. Assumes US economic activity is weakly exogenous with respect to Wisconsin. Source: Philadelphia Fed, NBER, author’s calculations.
In other words, Wisconsin is, with statistical significance, underperforming what would have been expected from historical experience. One argument is that Wisconsin’s statistical underperformance, which has intensified in recent months, is due to the dollar’s recent strength. This is a plausible interpretation for the last year ; however it fails to explain why underperformance was in evidence during the prior period of dollar weakness.
Fortunately for Wisconsin residents, our Governor Walker has assured us that The state of our state is strong!.