Purely as stimulus, a stimulus program attempts to increase economic activity and decrease unemployment. Those are linked but not identical goals, since the same level of economic activity can support different numbers of jobs depending on the average wage or salary of the jobs involved; you can make more $20,000 jobs for a billion dollars’ worth of stimulus than you can $100,000 jobs.
The stimulus value – the macroeconomic effect – of an expenditure or tax cut depends on (1) how quickly it gets paid out; (2) how much of the resulting additional income gets spent rather than saved or used to pay down debt; (3) how much of the additional income generated by the amount that is spent rather than saved is itself spent rather than saved (i.e., the multiplier); and (4) how much of it gets spent for domestic products rather than exports.
So claims like “spending on family planning doesn’t stimulate the economy” or “spending on re-sodding the National Mall doesn’t stimulate the economy” are simply nonsensical, suitable for recitation by knaves to fools. “Buy American” provisions help on #4, but only if we make the implausible assumption that other countries won’t make the same provisions in their stimulus packages.
In addition to its effect on current economic activity and unemployment, the money spent (or returned in tax cuts) under the stimulus plan will be spent on something, and those things may have different levels of current or future value: microeconomic effects. Helping people buy SUVs will stimulate the economy, but make us worse off in the future in both economic and environmental terms compared to helping people buy hybrids.
Finally, both the direct stimulus expenditures and the resulting changes in economic activity have distributional effects: we all care who gets how much, with Republicans wanting more of it to go to the rich and Democrats wanting more of it to go to the poor.
Got it? Macroeconomic effects, microeconomic effects, and distributional impacts. That’s all there is; there ain’t no more.