Attitudes towards Digitalization and Automatization in the Labor Market.

How to cite

Doerrenberg, Philipp; Melanie, Arntz; Blesse, Sebastian. 2019. "Attitudes towards Digitalization and Automatization in the Labor Market.." AEA RCT Registry. AEARCTR-0003888. https://doi.org/10.1257/rct.3888-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Philipp Doerrenberg* , Arntz Melanie , Sebastian Blesse

* Primary investigator

Registry ID
AEARCTR-0003888
DOI
https://doi.org/10.1257/rct.3888-1.0
First registered
Last updated
Status
In development
Trial dates
2019-02-18 – 2019-02-28
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/3888

Abstract

The implications of digitaization and automatization for the labor market and employment are heavily debated, both in the public and among academics. While some of the press coverage leaves the impression that the demand for human work will substantially decrease as a response to automatization and digitalization, some recent academic work finds evidence that automatization and digitalization increase overall human employment but have distributional consequences. In light of the public and academic debate, we study attitudes towards digitalization and automatization using a large-scall representative survey with experimental components in Germany and the US.

We survey many different attitudes, opinions, information and perceived employment threats in the context of recent developments in the labor market. We also ask for potential policy implications, attitudes towards redistribution and have a real behavioral outcome variable where we offer the actual possibility to donate money for charity. One aim of our study therefore is to gain a detailed picture of the perceptions in the context of the topic among the general public. We also include a randomized component into the survey where we manipulate the level of information about the topic. The information treatments confront participants with the results of a recent academic paper by Graetz and Michaels (2018, ReStat). This experimental component aims at studying whether biased information about the topic change attitudes and perceptions. An additional randomized component varies the order along which our questions are presented: we sometimes first ask general questions whereas sometimes we first ask specific questions about automatization and digitalization. This component helps to study if the general awareness of the topic changes partcioants' perceptions.

Intervention

We randomly split all survey participants into four groups:

Group 1: Neutral information; normal order of questions (control group)
Group 2: Neutral information; order of questions different than for group 1
Group 3: Information about employment effects of automatization (based on Graetz and Michaels (2018, ReStat)); order of questions as in group 2
Group 4: Information about distributional effects of automatization (based on Graetz and Michaels (2018, ReStat)); order of questions as in group 2

Experimental design

a) We include a randomized component into the survey where we manipulate the level of information about the topic. The information treatments confront participants with the results of a recent academic paper by Graetz and Michaels (2018, ReStat). This experimental component aims at studying whether biased information about the topic change attitudes and perceptions.

b) An additional randomized component varies the order along which our questions are presented: we sometimes first ask general questions whereas sometimes we first ask specific questions about automatization and digitalization. This component helps to study if the general awareness of the topic changes partcioants' perceptions.

Overall, we randomly split all survey participants into four groups:

Group 1: Neutral information; normal order of questions (control group)
Group 2: Neutral information; order of questions different than for group 1
Group 3: Information about employment effects of automatization (based on Graetz and Michaels (2018, ReStat)); order of questions as in group 2
Group 4: Information about distributional effects of automatization (based on Graetz and Michaels (2018, ReStat)); order of questions as in group 2

Design details

Randomization method
Randomization done by computer
Randomization unit
Individual survey participants
Clusters
5000
Observations
5000
Arms
Evenl split into four groups: about 1250 observations per randomized group.
Primary outcomes
We are potentially interested in the efefcts of the randomized treatments on any of the questions that we ask. In partucluar, we are interested in the effect on a set of variables asking about the personal and general consequences of automatization and the policy implications of these trends (including an actual behavioral outcome variable based on a charity possibility that we offer to participants)

Cited by

No citation data available for this registration yet.

Keywords