A preponderant challenge that affects research for poverty reduction and limits its potential value is the mismatch between the complex nature of the problem and the reductionist nature of what measures attempt to define and track it. Three examples illustrate the challenge. First, a growing number of so-called developing countries are achieving growth rates that (sustainably) surpass those of higher-income countries, which means that research should next investigate what policies can reduce inequality within developing countries and turn their economic growth into jobs for the benefit of all the members of their societies. Second, rapid urbanization, especially in Asia, means that monetary measures of poverty on which so much past research relied can no longer usefully differentiate between urban and rural contexts because the latter are less monetized. Third, a very different perspective must be shone on, and lessons drawn from, the new risks, shocks, and uncertainties that globalization has spawned (and which include price hikes impacting food). And so, new methodologies that measure multidimensional poverty must be honed: most likely, they must rest on a mix of (longitudinal) quantitative and qualitative research.
In the short run, there is Phillips curve which predicts negative relationship between inflation and unemployment; partly due to wage rigidity. Therefore, policy to stabilize inflation may end up raising unemployment.
In the long run, however, under neoclassical economics theory; when wage is assumed as flexible; supply will always equate demand of labor, i.e. no unemployment. Unemployment occurs under some policy to guarantee minimum wage.
As the best policy to reduce poverty is by providing job, I think you should also consider employment theory.