icaragua is a prominent agricultural country, such as the 28.1 % of the GDP, the 15.9 % of the total exports, and the 42.6 % of the national employment is given by agricultural sector. The mean features of small farmers are: a) They represent the 80 % as of total farmers, while they are owner of 24 % as of total land; b) They are 80 % men and 20 % are women, c) They have only 0.02 % a basic education; d) They have 46 % title deed, 16 % without title deed, 13 % in process of legalization, and the rest other form of possession (NIID, III CENAGRO: 2001). The importance of this study is focused in explain why does Nicaragua being an agricultural country then the workers were looking for other alternatives on nonfarm agricultural activities. This problem was reflected by the migration to town, or other neighbor Central American countries and the rural household need to generate wage and employ when the public policy measure was applied. N I. in section 4, and conclusion and discussion is showed in section 5. The concept ¨non-farm agricultural¨ was generated by rural farmers in secondary and tertiarysectors where RNFE and RNFW was employed and income indexed (Berdegué et al., 2000), other authors define it as derive of rural area which define the rural non-agricultural economy (RNFAE): activities and incomes. The RNAE is often defined as including all economic activities in rural areas except agriculture, livestock, hunting and fishing (Lanjouw and Lanjouw, 1997). More over ¨Non-Farm¨ is defined as being all those diverse activities associated with waged work or self-employment in work that is not agriculture but located in rural areas (David and Pearce, 2000). During period 1950 the 54 per cent was busy in agricultural activities from the rural sector of Latin America, however in 1990 only 25 per cent was in it (Milicevic, 2000). This was explained by both ruralurbanmigrations and framework change in rural labour market. The past investigations in some countries show that RINFA is a high and increase ratio of the total rural poor household in last decade (Berdegué et al., 2000). It is a strategies livelihood 2 . The both RNFE and RNFW are part of it. On the other hand, analysis of rural regions of the EU can point to issues of importance for the transitions economies. Outside Central Europe this studies in this field are now being undertaken, since it is recognized that in the longer term the development of the rural non-farm sector is critical factor in providing ruralemployment and income (Bleahuand Janowski, In the reviews of empirical studies we find that some studies were based in the concept of rural, nonfarm agricultural, non-farm income, and non-farm employment. Others authors explain the relation between rural employ and non-farm income, the mitigation process of rural poverty, of transformation farming and livestock sector, and transformation modern rural sector. Even they discussed the trend both employ and non-farm income. They also discussed the kind difference both employ and income non-farm. and in kind; the social relations and institutions that facilitate or constrain individual or family standards of living; and access to social and public services that contribute to the well-being of the individual or family." 2001; Breischopf and Schreider, 1999; Deichmann and Henderson, 2000;Chaplin, 2000;Sarris et al., 1999). In countries such as Romania, where agriculture is acting as a buffer against unemploymentand hidden unemployment is widespread and increasing (Da vis and Pearce, 2000), so RNAE is important for poverty reduction. # a) Binary dependent variable model In this class of models, authors discuss estimation methods for several qualitative and limited dependent variables models. Some software provides estimation routines for binary or ordered (probit, logit, gompit) censored or truncated (Tobit, etc.), and integer valued (count data) models. Standard introductory discussion for the models presented in this section may be found in Greene(1997), Johnston and DiNardo (1997), and Maddala (1983). Wooldridge (1996) provides an excellent reference for quasi-likelihood methods and count models. In this class of models , the dependent variable, may take on only two values-might be a dummy variable representing the occurrence of an event, or a choice between two alternatives. For example, you may be interested in modeling the employment status of each individual in your sample (whether employed or not). The individuals differ in age, educational attainment, race, marital status, and other observable characteristics, which can be denote as . The goal is to quantify the relationship between the individual characteristic and the probability of being employed. In the binary dependent variable model, the dependent variable, may take on only two values 0-1 might be a dummy variable representing the occurrence of an event (in our case is employment), or a choice between two alternatives: employ in agricultural activities or employ in nonfarm agricultural activities. Suppose that we model the probability of observing a value of one as: where F is a continuous, strictly increasing function that takes a real value and returns a valueranging from zero to one. The choice of the function F determines the type of binary model. Itfollows that: Given such a specification, we can estimate the parameters of this model using the method of maximum likelihood. The likelihood function is given by: The first order conditions for this likelihood are nonlinear so that obtaining parameter estimates requires an iterative solution. I use Eviews 5.1 that by default, it uses a second derivative method for iteration and computation of the covariance matrix of the parameter estimates. There are two alternative interpretations of this specification that are of interest. First, the binary model is often motivated as a latent variables specification. Suppose that there is an unobserved latent variable. whereis a random disturbance. Then the observed dependent variable is determined by whether * exceeds a threshold value: In this case, the threshold is set to zero, but the choice of a threshold value is irrelevant, so long as a constant term is included in . Then: where ' is the cumulative distribution function of ?. Common models include probit (standardnormal), logit, (logistic), and gompit (extreme value) specification for the F function. In principle, the coding of the two numerical values of y is not critical since each of the binary responses only represents an event. Nevertheless, Eviews require that I code y as zero-one variable. This restriction yields a number of advantages. For one, coding the variable in this fashion implies that y = 1. Pr ??(? = 1 ?Ý?"/? , ?) = 1 ? ??(?? ?Ý?"? ? ?)(1) Pr ??(? = 0 ?Ý?"/? , ?) = ??(?? ?Ý?"? ? ?)(2)?(?) = ??? ? log (1 ? ?Ý?"?(?? ? ?)) + (1 ? ??? )log (????? ?Ý?"? ? ??(3)??? * = ?Ý?"? ? ? + ? (4) ??? ??? x. ??? ??? ??? = ? 1 ?? ??? * > 0 0 ?? ??? * ? 0 (5) Pr ??(? = 1 ?Ý?"/? , ?) = Pr ??(? * > 0) = Pr ?Ý?"(? ? ? + ? > 0) = 1 ? ??? à°?" (? ?Ý?"? ? ?) (6) Global Journal of Management 16 2012 ear Y This model was used because the study is focused in the employment behavior. I was interested in modeling the employment status of each Working Economic Population (more than 10 year and less than 60 year). # III. # 2012 Year This convention provides us with a second interpretation of the binary specification as a conditional mean specification. It follows that we can write the binary model as a regression model: As Eviews requires code dependent variable, it is coding as a zero-one. One if the farm employs working economic population in agricultural activities, zero if the farm no employs it. In the other hand, there are two groups for coding independent variable. The first group is for wage and the second is for employ. The first it is coding as salary index, the calculation for is as follows: Where, is the monthly real wage index of each farm; is the weightier of either farm or nonfarm agricultural activity "K" and finally is the simple index for the farm activity "K". The weightier by each farm activity is getting of divide it between the total farm wages in a year. It is as follow: Where, is the participation of each farm activity in the total earnings; ;<=)(>) is the income of each farm activity "K"; and ?@?