How does the BigQuery command line tool "bq query" stop printing the sql query string for some jobs?

Question

Let's say the following file exists in the current directory:

# example1.sql

SELECT 1 AS data

When I run bq query --use_legacy_sql=false --format=csv < example1.sql I get the expected output:

Waiting on bqjob_r1cc6530a5317399f_0000017f2b966269_1 ... (0s) Current status: DONE   
data
1

But if I run the following query

# example2.sql

DECLARE useless_variable_1 STRING DEFAULT 'var_1';
DECLARE useless_variable_2 STRING DEFAULT 'var_2';

SELECT 1 AS data;

like this bq query --use_legacy_sql=false --format=csv < example2.sql I get the following output:

Waiting on bqjob_r501cc26a3d336b7c_0000017f2b97ceb1_1 ... (0s) Current status: DONE   
SELECT 1 AS data; -- at [6:1]
data
1

As you can see, adding variables caused bq query to print the contents of "example2.sql" to stdout, which is not what I want. How can I avoid printing the contents of the file to stdout and instead just return the result of the query like in the first example?