diff --git a/lms/djangoapps/class_dashboard/dashboard_data.py b/lms/djangoapps/class_dashboard/dashboard_data.py
index 255b8cefa2..6d059ac135 100644
--- a/lms/djangoapps/class_dashboard/dashboard_data.py
+++ b/lms/djangoapps/class_dashboard/dashboard_data.py
@@ -470,7 +470,7 @@ def get_students_opened_subsection(request, csv=False):
         # Subsection name is everything after 3rd space in tooltip
         filename = sanitize_filename(' '.join(tooltip.split(' ')[3:]))
 
-        header = [_("Name").encode('utf-8'), _("Username").encode('utf-8')]
+        header = [_("Name"), _("Username")]
         for student in students:
             results.append([student['student__profile__name'], student['student__username']])
 
@@ -528,9 +528,8 @@ def get_students_problem_grades(request, csv=False):
         tooltip = request.GET.get('tooltip')
         filename = sanitize_filename(tooltip[:tooltip.rfind(' - ')])
 
-        header = [_("Name").encode('utf-8'), _("Username").encode('utf-8'), _("Grade").encode('utf-8'), _("Percent").encode('utf-8')]
+        header = [_("Name"), _("Username"), _("Grade"), _("Percent")]
         for student in students:
-
             percent = 0
             if student['max_grade'] > 0:
                 percent = round(student['grade'] * 100 / student['max_grade'])
@@ -556,10 +555,13 @@ def post_metrics_data_csv(request):
 
     results = []
     if data_type == 'subsection':
-        header = [_("Section").encode('utf-8'), _("Subsection").encode('utf-8'), _("Opened by this number of students").encode('utf-8')]
+        header = [_("Section"), _("Subsection"), _("Opened by this number of students")]
         filename = sanitize_filename(_('subsections') + '_' + course_id)
     elif data_type == 'problem':
-        header = [_("Section").encode('utf-8'), _("Problem").encode('utf-8'), _("Name").encode('utf-8'), _("Count of Students").encode('utf-8'), _("Percent of Students").encode('utf-8'), _("Score").encode('utf-8')]
+        header = [
+            _("Section"), _("Problem"), _("Name"), _("Count of Students"),
+            _("Percent of Students"), _("Score"),
+        ]
         filename = sanitize_filename(_('problems') + '_' + course_id)
 
     for index, section in enumerate(sections):
diff --git a/lms/djangoapps/instructor/views/api.py b/lms/djangoapps/instructor/views/api.py
index a21db446bc..aef690b235 100644
--- a/lms/djangoapps/instructor/views/api.py
+++ b/lms/djangoapps/instructor/views/api.py
@@ -1112,7 +1112,7 @@ def get_sale_order_records(request, course_id):  # pylint: disable=unused-argume
     db_columns = [x[0] for x in query_features]
     csv_columns = [x[1] for x in query_features]
     sale_data = instructor_analytics.basic.sale_order_record_features(course_id, db_columns)
-    header, datarows = instructor_analytics.csvs.format_dictlist(sale_data, db_columns)  # pylint: disable=unused-variable
+    __, datarows = instructor_analytics.csvs.format_dictlist(sale_data, db_columns)
     return instructor_analytics.csvs.create_csv_response("e-commerce_sale_order_records.csv", csv_columns, datarows)
 
 
diff --git a/lms/djangoapps/instructor_analytics/csvs.py b/lms/djangoapps/instructor_analytics/csvs.py
index c03a6b7456..2b8da615f0 100644
--- a/lms/djangoapps/instructor_analytics/csvs.py
+++ b/lms/djangoapps/instructor_analytics/csvs.py
@@ -14,10 +14,13 @@ def create_csv_response(filename, header, datarows):
 
     header   e.g. ['Name', 'Email']
     datarows e.g. [['Jim', 'jim@edy.org'], ['Jake', 'jake@edy.org'], ...]
+
+    The data in `header` and `datarows` must be either Unicode strings,
+    or ASCII-only bytestrings.
+
     """
     response = HttpResponse(content_type='text/csv')
-    response['Content-Disposition'] = 'attachment; filename={0}'\
-        .format(filename)
+    response['Content-Disposition'] = 'attachment; filename={0}'.format(filename)
     csvwriter = csv.writer(
         response,
         dialect='excel',
@@ -30,6 +33,7 @@ def create_csv_response(filename, header, datarows):
     for datarow in datarows:
         encoded_row = [unicode(s).encode('utf-8') for s in datarow]
         csvwriter.writerow(encoded_row)
+
     return response