台灣農民與非農民之數位落差分析

台灣為電腦科技大國，台灣網路普及率高達9成，是個高度的資訊化社會，然而數位落差實質上正悄悄的擴大於農業社會當中。台灣政府農業單位近年重視農業資訊化，建立近百個農業資訊相關的系統平台，如台灣農產品生產追溯系統、產銷履歷農產品資訊網、農業科技產業情報站等。在建置此資訊平台的同時，盤點農業資訊化程度是個值得了解的課題。本研究以是否使用即時通訊軟體 (Instant Messaging)作為指標，透過假設檢定估計農民與非農民數位落差是否存在，再藉由Probit模型估計那些因素影響即時通訊軟體使用狀況，最後利用Blinder-Oaxaca Decomposition模型分析性別、年齡、月收入以及教育程度影響農民與非農民數位落差程度，分析出3G時代(2009年至2012年)與4G時代(2015年至2018年)數位落差變動狀況，藉此了解農民與非農民之間的數位落差。估計結果顯示：無論3G時代或4G時代，農民與非農民皆存在數位落差；年齡與使用即時通訊軟體的機率呈現負相關，而月收入與教育水準則呈現正相關。透過Blinder-Oaxaca Decomposition模型分析結果則可發現，其中稟賦效果為可解釋部分都超過70%以上；若比較性別，3G與4G時代男性數位落差皆比女性小。在三個不同的年齡層(15-44歲、45-64歲、65歲以上)當中，3G時代與4G時代相同影響稟賦效果第一名者皆為教育程度，分別為61.42%、43.06%、48.21%，尤其在15-44歲年齡層中影響最大，高達六成。由此研究可知，由3G時代進入4G時代後，農民與非農民數位落差依舊存在。政府在制定政策與資源分配上應著重在教育層面，並且推廣更多年輕人加入農業。

關鍵字

數位落差；即時通訊軟體； Probit模型； Blinder-Oaxaca Decomposition模型

並列摘要

Taiwan is a highly developed information society, with over 90% people as Internet users. However, digital divide becomes larger and larger between agricultural industry and non-agricultural industry. To promote the informatization of the agricultural industry, Taiwan government established many computing platforms, such as Taiwan agricultural products production traceability system, Taiwan agriculture and food traceability system, agricultural technology industry information platform, etc. Besides establishing all these platforms, it is also an important issue to evaluate the level of our agricultural informatization. This study uses the instant messaging usage as the measurement of digital capability to address several issues. First, we investigate whether there exists digital divide between agricultural and non-agricultural employees. We then examine the digital divide among different socio-demographic groups. Data used in the empirical analysis consists of people aged 15 and over drawn from the Taiwan Digital Divide Survey from 3G era (2009 to 2012) to 4G era (2015 to 2018). The empirical analysis was conducted in a two-stage framework. The first stage analysis estimates instant messaging using the Probit Model, and the second stage analysis decomposes the agricultural-nonagricultural digital divide using Blinder-Oaxaca Decomposition Model. Results show that there was a digital divide between agricultural and non-agricultural employees, regardless of the 3G or 4G era. A negative association is found between instant messaging usage and age. In contrast, instant messaging usage is positively correlated with monthly income and educational level. Probability of instant messaging usage is higher for male in both 3G and 4G era. In the results of the decomposition analysis, the explained part in digital divide accounted for about over 70% of the total digital inequality. Also, the digital divide is mainly driven by the differences in the differences in education level, especially in the 15- to 44-year-old generation (61.42%). In conclusion, as Taiwan joined the 4G era in recent years, there is still a digital divide between agricultural and non-agricultural employees. To overcome the digital divide, the government should put emphasis on education when making policies and allocating resources. The government should also recruit more young people to agricultural industry.