APPLICATION OF THE P/F RATIO METHOD IN ESTIMATING FERTILITY LEVELS IN LESOTHO By THANDIE KEROMAMANG HLABANA Submitted in Partial Fulfilment of the Research Requirements for the Degree of Masters in Development

Inadequate demographic data in Lesotho inhibits demographic research. Nonetheless, indirect demographic techniques have proven to be useful tools in the developing world, as their application to census and survey data has greatly expanded knowledge of the demographic situation in data deficient countries (Brass, 1996). The different techniques are based on specific assumptions and robustness of available data, thus deserves caution in application. Failure to adhere to these methodological specifications results in generation of more errors (Feeney, 1996). The impetus of this research was to assess the applicability of the P/F ratio method in estimating recent fertility levels in Lesotho. In particular, the data was evaluated to verify the following P/F ratio assumptions; (1) constant fertility; (2) accurate reporting of fertility by younger women; and (3) correct age pattern of fertility. In order to obtain optimal fertility estimates, the research undertook extensive data assessment, and corrections where possible, of individual variables employed in the P/F ratio method. In line with previous studies, the magnitude and pattern of the P/F ratios represented strong evidence of fertility decline in Lesotho. This evidence rendered the Brass P/F ratio method inappropriate for estimation of recent fertility levels in the country. Therefore, this research presents the Relational Gompertz model faring better in indirectly estimating fertility levels in Lesotho. Not undermining the Bureau of Statistics, the current study challenges the Bureau's estimates, and declares own estimates as more likely precise estimates of recent fertility levels in Lesotho when using the P/F ratio method. This assertion is grounded on the basis that compared to the Bureau, the study undertook and presented detailed data evaluation and adjustments, as well as adhering to the P/F ratio methodological assumptions. Nonetheless, the research also concludes that indirect techniques do not necessarily provide an utopia to demographic estimation in poor data countries. Even when the robust measures were employed, the quality of the 1996 data yielded implausible estimates as the method could not account for the degree of unreported births. This calls for caution during data collection and processing in order to minimise the reporting errors.