This is

Very nice. I do object to one word " extraordinary " in " absent such an extraordinary need to boost government". My now familiar argument is that, given the equation, if it is not good policy to raise government spending, then it is good policy to cut government spending. I now suggest a parameter for the relative social value of goverment and private spending call it nu so welfare = Y + (nu-1)G. If nu is 1+ zeta, then G is right when mu =0. How low does it have to be for optimal delta G in a depressed economy to be negative ? Clearly in the first not cost benefit case, 0 is high enough. For mu = 1, the only good terms left are mulitples of eta -- the only good effect is due to hysteresis. I think equation 12 becomes arounf 6.5% > r ( doing all arithmetic in my head so prob 0 I got it all right).

Also how about pay back in 30 years with constant increase in T. The advantage is, for a small increase in G, you have all relevant r. No need to forecast as Treasury can sell 30 year TIPS. Of course for a non tiny increase in G you still need to guess what r will be once policy announced, but will no all you need to know before it's too late (policy sell at some minumum price all the bonds you can up to a ceiling, then spend the proceeds). Worth the arithmetic bother ?

## No comments:

## Post a Comment