000 | 04918nam a22004217a 4500 | ||
---|---|---|---|
001 | G77113 | ||
003 | MX-TxCIM | ||
005 | 20211006080815.0 | ||
008 | 121211s ||||f| 0 p|p||0|| | | ||
020 | _a970-648-076-5 | ||
040 | _aMX-TxCIM | ||
072 | 0 | _aA50 | |
072 | 0 | _aE14 | |
082 | 0 | 4 |
_a338.91 _bWAT |
100 | 1 |
_aHolloway, G. _uInternational conference on impacts of agricultural research and development: Why has impact assessment research not made more of a difference? |
|
110 | 0 | _aCentro Internacional de Mejoramiento de Maiz y Trigo (CIMMYT), Mexico DF (Mexico) | |
111 | 2 |
_aInternational Conference on Impacts of Agricultural Research and Development _cSan José (Costa Rica) _d4-7 Feb 2002 |
|
245 | 0 | 0 |
_aThe simple econometrics of impact assessment: _b theory with an application to milk-market participation in the ethiopian highlands |
260 |
_aMexico, DF (Mexico) _bCIMMYT : _c2003 |
||
300 | _ap. 56 | ||
340 | _aPrinted | ||
500 | _aAbstract only | ||
520 | _aThis paper presents novel econometric techniques for enhancing the robustness of impact assessments in the context of econometric investigations of research productivity, institutional innovation or poverty alleviation. The methodology has broad application to time series or cross-section data. However, it is especially useful when applied to household panel data (the focus of this study) when adoption and participation issues are in question. In these circumstances, one problem that frequently arises is a policy choice between alternative means of achieving stated objectives. Because the impacts of various policies are often predicated on econometric results, policy options are associated with a probability distribution about their impacts. When these distributions emanate from different parametric families, a problem of comparison arises. For example, the distributions of the precise impacts of two altemative policies may have different locations (as indicated by their respective means estimates) and may also have different scales (as indicated by their respective variance estimates).Even when the distribution of one policy stochastically dominates another, a problem of ranking arises when budgetary considerations limit resources. In this paper, we demonstrate how measurement problems of stochastic impact assessment can be resolved tractably, intuitively and robustly as an outcrop of most conventional econometric modeling exercises. The methodology exploits recent advances in Markov Chain Monte Carlo (MCMC) methods. MCMC methods are a collection of iterative techniques for estimating models in which discrete-choice, truncated or censored-regression formulations are at issue. This general class of models appears frequently in empirical investigations of agricultural productivity, equity, poverty, social health, and nutrition, especially where the traditional household-production model is the modus operandi. Surprisingly, there have been no applications of these powerful techniques to impact assessment. Hence, the study represents the first of its kind. In our application, we present empirical results related to an emerging market in the Ethiopian highlands. We explore the measurement of various policies targeted toward promoting participation among formerly subsistence milk producers at two peri-urban sites close to the capital city, Addis Ababa. In this context, the policy questions include the identification of the covariates and the levels that are required to effect entry by non-participants in a market; in other words, measures of the inputs needed to create and sustain a market. This application is important in the context of developing economies, where the density of non-participants is high and is considered to be the main impediment to economic development. Our main objective was to estimate the covariate levels that prompt market entry of non-participating households. The strategy was very simple. First, we estimated each non-participant's distance from the market in terms of the (latent) level of marketable surplus that the household would wish to produce, given its covariates. These are the augmenting data. By definition, they are negative real numbers. Subsequently, we computed the levels of the covariates that make these negative numbers become positive, signifying positive marketable surplus and, presumably, entry into the market. | ||
546 | _aEnglish | ||
591 | _a0310|R01CIMPU|AGRIS 0301|AL-Economics Program | ||
593 | _aJuan Carlos Mendieta | ||
595 | _aCPC | ||
650 | 1 | 0 | _aEconometric models |
650 | 1 | 0 | _aEconometrics |
650 | 1 | 0 | _aEconomic resources |
650 | 1 | 0 | _aPrice policies |
650 | 1 | 0 |
_aProductivity _gAGROVOC _91756 |
653 | 0 | _aCIMMYT | |
650 | 1 | 7 |
_aAgricultural research _gAGROVOC _2 _91006 |
700 | 1 |
_aEhui, S., _ecoaut. |
|
700 | 1 |
_9960 _aWatson, D.J. _gResearch & Partnership Program _8INT3479 _eed. |
|
942 | _cPRO | ||
999 |
_c6862 _d6862 |