Interesting Implied Macro Impacts Of Representative Ryan’s Roadmap

I’ve read and re-read the Heritage Foundation’s analysis of how the projections for the Ryan plan were developed. I’m sure it’s my own failing, but I still don’t quite understand what is going on. And this is after Heritage took down their original documentation that indicated unemployment would eventually hit 2.8%.[0]

(Here is National Journal’s take on the original Heritage analysis.)

Even ignoring the unemployment number (which seems to have moved a bit, although not reported in the document), I thought it worthwhile to mention the other oddities of the report.

First, it is important to note that the simulation forecasts relative to the CBO alternative fiscal scenario, rather than extended baseline, as would typically be the case. Obviously, this makes the Ryan plan “look better” in terms of budget deficits and (given Heritage’s modelling approach incorporating substantial supply side effects) in terms of growth.

Second, it is very interesting to take a look at the forecasts. For GDP (Figure 1), the forecasts imply a noticeable increase, amounting to a 2.4% higher GDP (in log terms, relative to baseline) by 2021. Perhaps reflecting the assumptions built into the model, despite reduced effective personal tax rates (see Appendix 3 tables), personal tax receipts are higher (Figure 2).


Figure 1: GDP, bn Ch.2005 actual and CBO projected (blue line); and simulated under Ryan plan (red line). Source: CBO, Budget and Economic Outlook (January 2011), and Heritage Foundation, Appendix 3: Simulation Results.


Figure 2: Baseline personal income tax receipts baseline (blue line); and simulated under Ryan plan (red line). Source: Heritage Foundation, Appendix 3: Simulation Results.

I suspect (but do not know for certain) that this counterintuitive (to me) result can be understood if one reads this paragraph in the Heritage document (retrieved as of 4/6 8pm Pacific, version updated as of 4/6 11am):

Average Effective Personal Tax Rates. The add factor on the average effective federal personal income tax rate was changed by the percentage change from the baseline estimated in the microsimulation model. Adjusting the add factor allows for the dynamic indirect effects could continue to influence the average effective tax rate. The simulation was solved in stages. The final stage endogenously re-estimated the add factor on the average effective rate in order to target the percentage of revenue to GDP outlined in the House Budget Resolution.

Given my research interests, I found of particular interest the projections for the trade balance and foreign assets in the U.S. These are shown in Figures 3 and 4.


Figure 3: Net exports baseline (bn Ch.2005$ (blue line); and simulated under Ryan plan (red line). Source: Heritage Foundation, Appendix 3: Simulation Results.


Figure 4: Foreign assets in the U.S., bn$ at current cost, baseline (blue line); and simulated under Ryan plan (red line). Source: Heritage Foundation, Appendix 3: Simulation Results.

Figure 4 is hard to interpret; it indicates foreign holdings of US assets rise. Since the budget deficit is supposed to be smaller over most of the period of analysis, it’s hard to think this is higher foreign acquisition of Treasurys. On the other hand, this makes sense (kind of) given the larger net export deficit implied by Figure 3. I say “kind of” because the nominal current account should be linked to the current cost U.S. net international investment position, and not necessarily linked to a gross liability stock. Without more information regarding the innards of the model, and other assumptions, it’s difficult to determine why the foreign asset result occurs (and whether it is a consequential result).

In addition, Figure 3 is really hard for me to understand. Certainly, from a demand side perspective, the contractionary impulses arising from re-allocating health care burdens directly onto seniors and cutting government discretionary spending would be contractionary. So the source of this deteriorated trade balance must be attributable to some supply side effects. The rise in personal tax receipts (relative to baseline) despite a cut in the effective tax rate (see Appendix 3 tables) is consistent with the view that the expansionary effect emanates from the supply side. It might also come about from the increased amount of capital goods imports that arises from higher investment in plant and equipment.

This impression is buttressed by this passage, regarding how reduction in Federal debt levels work their way through the economy (page 13). 

Cost of Capital. The cost of capital changes are affected directly and indirectly by the dynamic effects. (1) Lower corporate tax rates reduces the value of the interest rate deduction, which can put upward pressure on the cost of capital. (2) Lower corporate tax rates, though increase the after-tax rate of return to capital, which puts downward pressure on its cost. (3) Lower government spending decreases the demand for borrowed funds, which puts downward pressure on the cost of funds. (4) labour and capital trade-offs as labour supply increases also plays an indirect role. These effects were all allowed to operate and then an adjustment was made to the federal funds rate, 10-year treasury rate, and corporate triple-A bond rate for the estimated percentage change in government debt. The model was then re-solved with this adjustment.25 

Private Investment. Economic studies repeatedly find that government debt crowds-out private investment although the degree to which it does so can be debated.26 The structure of the model does not allow for this direct feedback between government spending and private investment variables. Therefore, the add factors on private investment variables were also adjusted to reflect percentage changes in publicly held debt. This can also put upward pressure on the cost of capital (thus helping the model balance the demand and supply effects on the cost of capital). [emphasis added — mdc]

I believe a reduction in prospective net debt-to-GDP levels does have an impact on interest rates. [1]. The resulting effect on investment is a subject of considerable debate.

Here is my comment on my previous interchange with a Heritage Foundation analysis, based at least in some part on the IHS Global Insight model.

Since I’m not sure what’s going on in the Heritage CDA simulations (even the head economist at IHS Global Insight, which the simulations are supposed to be partly based upon, can’t quite figure out what is going on [2]), I’m going to forego the Heritage forecasts, and fall back on assessments from the CBO. Note, the CBO does not do “dynamic scoring” — incorporating feedback effects from budgetary measures into macro outcomes. (More on dynamic scoring: [3] [4]

As shown in Table 1, the Ryan plan actually yields a larger debt-to-GDP ratio than the extended baseline by 2022 (but less than the CBO Alternative Fiscal Scenario). Clearly, as we pass the transition phase, the gap between the debt-to-GDP trajectory widens, by virtue of cutting benefits by fiat.


Table 1 from CBO.

The impact on the burden of health care costs (which actually rise faster under the Ryan plan) is shown in Figure 1 from the CBO letter.


Figure 1 from CBO.

Update, 1:40pm Pacific: Paul Krugman scooped me by a day in noting that the Heritage CDA analysis of the Ryan plan invokes supply side effects of implausibly large proportions. For my quantitative view of “fantasy scenarios” such as these, see this post.